Skip to main content

Online Safety Bill

Volume 829: debated on Thursday 27 April 2023

Committee (3rd Day) (Continued)

Clause 8: Illegal content risk assessment duties

Amendment 16

Moved by

16: Clause 8, page 7, line 16, after “governance,” insert “terms of service,”

Member’s explanatory statement

This amendment makes clear that “design and operation of a service” includes its terms of service.

My Lords, this group of amendments concerns terms of service. All the amendments either have the phrase “terms of service” in them or imply that we wish to see more use of the phrase in the Bill, and seek to try to tidy up some of the other bits around that which have crept into the Bill.

Why are we doing that? Rather late in the day, terms of service has suddenly become a key fulcrum, under which much of the operations of the activity relating to people’s usage of social media and service functions on the internet will be expressed in relation to how they view the material coming to them. With the loss of the adult “legal but harmful” provisions, we also lost quite a considerable amount of what would have been primary legislation, which no doubt would have been backed up by codes of practice. The situation we are left with, and which we need to look at very closely, is the triple shield at the heart of the new obligations on companies, and, in particular, on their terms of service. That is set out primarily in Clauses 64, 65, 66 and 67, and is a subject to which my amendments largely refer.

Users of the services would be more confident that the Government have got their focus on terms of service right, if they actually said what should be said on the tin, as the expression goes. If it is the case that something in a terms of service was so written and implemented so that material which should be taken down was indeed taken down, these would become reliable methods of judging whether or not the service is the one people want to have, and the free market would be seen to be working to empower people to make their own decisions about what level of risk they can assume by using a service. That is a major change from the way the Bill was originally envisaged. Because this was done late, we have one or two of the matters to which I have referred already, which means that the amendments focus on changing what is currently in the Bill.

It is also true that the changes were not consulted upon; I do not recall there being any document from government about whether this was a good way forward. The changes were certainly not considered by the Joint Committee, of which several of those present were members—we did not discuss it in the Joint Committee and made no recommendation on it. The level of scrutiny we have enjoyed on the Bill has been absent in this area. The right reverend Prelate the Bishop of Oxford will speak shortly to amendments about terms of service, and we will be able to come back to it. I think it would have been appropriate had the earlier amendment in the name of the noble Lord, Lord Pickles, been in this group because the issue was the terms of service, even though it had many other elements that were important and that we did discuss.

The main focus of my speech is that the Government have not managed to link this new idea of terms of service and the responsibilities that will flow from that to the rest of the Bill. It does not seem to fit into the overall architecture. For example, it is not a design feature, and does not seem to work through in that way. This is a largely self-contained series of clauses. We are trying to ask some of the world’s largest companies, on behalf of the people who use them, to do things on an almost contractual basis. Terms of service are not a contract that you sign up to, but you certainly click something—or occasionally click it, if you remember to—by which you consent to the company operating in a particular set of ways. In a sense, that is a contract, but is it really a contract? At the heart of that contract between companies and users is whether the terms of service are well captured in the way the Bill is organised. I think there are gaps.

The Bill does have something that we welcome and want to hold on to, which is that the process under which the risks are assessed and decisions taken about how companies operate and how Ofcom relates to those decisions is about the design and operation of the service—both the design and the operation, something that the noble Baroness, Lady Kidron, is very keen to emphasise at all times. It all starts and ends with design, and the operation is a consequence of design choices. Other noble Baronesses have mentioned in the debate that small companies get it right and so, when they grow, can be confident that what they are doing is something that is worth doing. Design, and operating that design to make a service, is really important. Are terms of service part of that or are they different, and does it matter? It seems to me that they are downstream from the design: something can be designed and then have terms of service that were not really part of the original process. What is happening here?

My Amendments 16, 21, 66DA, 75 and 197 would ensure that the terms of service are included within the list of matters that constitute “design and operation” of the service at each point that it occurs. I have had to go right through the Bill to add it in certain areas—in a rather irritating way, I am sure, for the Bill team—because sometimes we find that what I think should be a term of service is actually described as something else, such as a “a publicly available statement”, whatever that is. It would be an advantage if we went through it again and defined terms of service and made sure that that was what we were talking about.

Amendments 70 to 72, 79 to 81 and 174 seek to help the Government and their officials with tidying up the drafting, which probably has not been scrutinised enough to pick up these issues. It may not matter, at the end of the day, but what is in the Bill is going to be law and we may as well try to get it right as best we can. I am sure the Minister will say we really do not need to worry about this because it is all about risks and outcomes, and if a company does not protect children or has illegal content, or the user-empowerment duties—the toggling—do not work, Ofcom will find a way of driving the company to sort it out. What does that mean in practice? Does it mean that Ofcom has a role in defining what terms of service are? It is not in the Bill and may not reach the Bill, but it is something that will be a bit of problem if we do not resolve what we mean by it, even if it is not by changing the legislation.

If the Minister were to disagree with my approach, it would be quite nice to have it said at the Dispatch Box so that we can look at that. The key question is: are terms of service an integral part of the design and operation of a service and, if so, can we extend the term to make sure that all aspects of the services people consume are covered by adequate and effective terms of service? There is probably going to be division in the way we approach this because, clearly, whether they are terms of service or have another name, the actual enforcement of illegal and children’s duties will be effected by Ofcom, irrespective of the wording of the Bill—I do not want to question that. However, there is obviously an overlap into questions about adults and others who are affected by the terms of service. If you cannot identify what the terms of service say in relation to something you might not wish to receive because the terms of service are imprecise, how on earth are you going to operate the services, the toggles and things, around it? If you look at that and accept there will be pressure within the market to get these terms of service right, there will be a lot of dialogue with Ofcom. I accept that all that will happen, but it would be good if the position of the terms of service was clarified in the Bill before it becomes law and that Ofcom’s powers in relation to those are clarified—do they or do they not have the chance to review terms of service if they turn out to be ineffective in practice? If that is the case, how are we going to see this work out in practice in terms of what people will be able to do about it, either through redress or by taking the issue to court? I beg to move.

I support these amendments, which were set out wonderfully by the noble Lord, Lord Stevenson. I want to raise a point made on Tuesday when the noble Baroness, Lady Merron, said that only 3% of people read terms of service and I said that 98% of people do not read them, so one of us is wrong, but I think the direction of travel is clear. She also used a very interesting phrase about prominence, and I want to use this opportunity to ask the Minister whether there is some lever whereby Ofcom can insist on prominence for certain sorts of material—a hierarchy of information, if you like—because these are really important pieces of information, buried in the wrong place so that even 2% or 3% of people may not find them.

My Lords, I am very pleased that the noble Lord, Lord Stevenson, has given us the opportunity to talk about terms of service, and I will make three points again, in a shorter intervention than on the previous group.

First, terms of service are critical as the impact of terms of service will generally be much greater in terms of the amount of intervention that occurs on content than it will ever be under the law. Terms of service create, in effect, a body of private law for a community, and they are nearly always a superset of the public law—indeed, it is very common for the first items of a terms of service to say, “You must not do anything illegal”. This raises the interesting question of “illegal where?”—what it generally means is that you must not do anything illegal in the jurisdiction in which the service provider is established. The terms of service will say, “Do not do anything illegal”, and then they will give a whole list of other things, as well as illegality, that you cannot do on the platform, and I think this is right because they have different characteristics.

Secondly, to back up the point made by the noble Baroness, Lady Kidron, we need to be realistic that no one will ever read all the terms of service of the services that they use. There was a study that looked at how long it would take to read the terms of service on a typical mobile phone—I think it is around 10 days; given that they get updated most years, are any of us going to spend 10 days a year reading the terms of service?

We like our real-world analogues: we read all of the law, but none of the people out there read all of the laws of the land unless and until they have a problem, at which point they do read them. Terms of service are very similar in that people are not going to read them and we should not expect people to read them unless and until they have a problem that requires them to do so. I do not mean that as a counsel of despair, but we have to be realistic about what we are expecting people to do.

Thirdly, the Bill is going to make terms of service longer, and we need to get over that. The challenge is always that you want your terms of service to be comprehensive and easy for users, and as we move in the Bill towards making terms of service more actionable—which we are doing because the Bill says that Ofcom will be able to say, “Did you apply your terms of service properly?”—the lawyers for the platforms are going to be saying “What have we missed out?” and “If there is anything we have missed out, we have to go and stick it in there because now we are going to have a regulator breathing down our neck, checking whether or not we have done what we say”.

We should be realistic that we are asking companies to be entirely comprehensive and transparent, and in general that will mean making their terms of service longer. Again, this is not a complete counsel of despair. We can follow Mark Twain’s advice:

“I didn’t have time to write you a short letter, so I wrote you a long one”

and invest the time. That is what we can do to try to make terms of service shorter, rather than just saying to lawyers, “We will pay you by the word, and the more words there are, the happier we are”. But, again, we should be realistic: if it is comprehensive, it is going to be long; there is no way to avoid that.

The noble Lord, Lord Stevenson, asked whether or not it is a contract—that is an interesting question, certainly for the US providers. In the US the regulation, such that there is, is done largely by the Federal Trade Commission, and the concept is whether or not services are engaged in unfair or deceptive practices. An unfair or deceptive practice is not doing what you said you would do in the terms of service or a critical document of that nature.

Interestingly, all the incentive in the US is to be as vague as possible because if you have not said that you will do things, you cannot be hauled in front of the FTC. The EU generally creatives incentives to be as comprehensive as possible, and I was involved in a number of cases where the company I worked for was taken to court and forced to add in more text because the US text was seen as too skeletal—that is a familiar debate to us here, whether we like things to be skeletal or for everything to be filled in.

So we need to be cognisant of that as we build terms of service into the Bill. This is not an argument against the amendments, but rather to say that as we do this, we need to be clear that we may be pulling in opposite directions. They need to be comprehensive, yet easy to use. “We are going to hold you accountable in the US; therefore, you should be vague; but we are also going to hold you accountable in the UK if you are too vague”—where is the right point of specificity and vagueness?

Having said that, it is really important that we focus on this because from a user’s point of view you are far more likely to come across an issue with the terms than an issue with the law—this is great, because most people in this country are law-abiding and not seeking to break the law.

The final point is that sometimes there is a tendency to think that everyone should have uniform terms of service. I can see the argument for a baseline, but in a vibrant market there is a strong case to say that we should celebrate where they are different, and there are communities that are different. For example, if you have a service that targets young people you might want to prohibit swearing; whereas, for example, it would be completely inappropriate to prohibit swearing in a vibrant political community for adults only. There are lots of areas where people understand that the context is different. For example, there are places where nudity—not pornography—is okay, and places where it is not.

So having different terms of service for different types of service is healthy, but I also think that Ofcom making sure that people do what they say they do is a reasonably healthy development, as long as we recognise and accept the consequences of that.

My Lords, I am grateful for this short and focused debate, which has been helpful, and for the points made by the noble Lords, Lord Stevenson and Lord Allan, and the noble Baroness, Lady Kidron. I think we all share the same objective: ensuring that terms of service promote accountability and transparency, and empower users.

One of the Bill’s key objectives is to ensure that the terms of service of user-to-user platforms are suitable and effective. Under the Bill, companies will be required both to set out clearly how they will tackle illegal content and protect children and to ensure that their terms of service are properly enforced. The additional transparency and accountability duties on category 1 services will further ensure that users know what to expect on the largest platforms. This will put an end to these services arbitrarily removing content or, conversely, failing to remove content that they profess to prohibit.

The Bill will also ensure that search services are clear to their users about how they are complying with their adult and child safety duties under this new law. Given the very different way in which search services operate, however, this will be achieved through a publicly available statement rather than through terms of service. The two are meant distinctly.

Noble Lords are right to point to the question of intelligibility. It struck me that, if it takes 10 days to read terms of service, perhaps we should have a race during the 10 days allotted to this Committee stage to see which is quicker—but I take the point. The noble Lord, Lord Allan, is also right that the further requirements imposed through this Bill will only add to that.

The noble Baroness, Lady Kidron, asked a fair question about what “accessibility” means. The Bill requires all platforms’ terms of service for illegal content and child safety duties to be clear and accessible. Ofcom will provide guidance on what that means, including ensuring that they are suitably prominent. The same applies to terms of service for category 1 services relating to content moderation.

I will focus first on Amendments 16, 21, 66DA, 75 and 197, which seek to ensure that both Ofcom and platforms consider the risks associated with platforms’ terms of service with regard to the illegal content and child safety duties in the Bill. We do not think that these amendments are needed. User-to-user services will already be required to assess the risks regarding their terms of service for illegal content. Clause 8 requires companies to assess the “design and operation” of a service in relation to illegal content. As terms of service are integral to how a service operates, they would be covered by this provision. Similarly, Clause 10 sets out that companies likely to be accessed by children will be required to assess the “design and operation” of a service as part of their child risk assessments, which would include the extent to which their terms of service may reduce or increase the risk of harm to children.

In addition to those risk assessment duties, the safety duties will require companies to take proportionate measures effectively to manage and mitigate the risk of harm to people whom they have identified through risk assessments. This will include making changes to their terms of service, if appropriate. The Bill does not impose duties on search services relating to terms of service, as search services’ terms of service play a less important role in determining how users can engage on a platform. I will explain this point further when responding to specific amendments relating to search services but I can assure the noble Lord, Lord Stevenson, that search services will have comprehensive duties to understand and mitigate how the design and operation of their service affects risk.

Amendment 197 would require Ofcom to assess how platforms’ terms of service affect the risk of harm to people that the sector presents. While I agree that this is an important risk factor which Ofcom must consider, it is already provided for in Clause 89, which requires Ofcom to undertake an assessment of risk across regulated services. That requires Ofcom to consider which characteristics of regulated services give rise to harm. Given how integral terms of service are to how many technology companies function, Ofcom will necessarily consider the risk associated with terms of service when undertaking that risk assessment.

However, elevating terms of service above other systems and processes, as mentioned in Clause 89, would imply that Ofcom needs to take account of the risk of harm on the regulated service, more than it needs to do so for other safety-by-design systems and processes or for content moderation processes, for instance. That may not be suitable, particularly as the service delivery methods will inevitably change over time. Instead, Clause 89 has been written to give Ofcom scope to organise its risk assessment, risk register and risk profiles as it thinks suitable. That is appropriate, given that it is best placed to develop detailed knowledge of the matters in question as they evolve over time.

Amendments 70, 71, 72, 79, 80, 81, 174 and 302 seek to replace the Bill’s references to publicly available statements, in relation to search services, with terms of service. This would mean that search services would have to publish how they are complying with their illegal content and child protection duties in terms of service rather than in publicly available statements. I appreciate the spirit in which the noble Lord has tabled and introduced these amendments. However, they do not consider the very different ways in which search services operate.

User-to-user services’ terms of service fulfil a very specific purpose. They govern a user’s behaviour on the service and set rules on what a user is allowed to post and how they can interact with others. If a user breaks these terms, a service can block his or her access or remove his or her content. Under the status quo, users have very few mechanisms by which to hold user-to-user platforms accountable to these terms, meaning that users can arbitrarily see their content removed with few or no avenues for redress. Equally, a user may choose to use a service because its terms and conditions lead them to believe that certain types of content are prohibited while in practice the company does not enforce the relevant terms.

The Bill’s duties relating to user-to-user services’ terms of service seek to redress this imbalance. They will ensure that people know what to expect on a platform and enable them to hold platforms accountable. In contrast, users of search services do not create content or interact with other users. Users can search for anything without restriction from the search service provider, although a search term may not always return results. It is therefore not necessary to provide detailed information on what a user can and cannot do on a search service. The existing duties on such services will ensure that search engines are clear to users about how they are complying with their safety duties. The Bill will require search services to set out how they are fulfilling them, in publicly available statements. Their actions must meet the standards set by Ofcom. Using these statements will ensure that search services are as transparent as user-to-user services about how they are complying with their safety duties.

The noble Lord’s Amendment 174 also seeks to expand the transparency reporting requirements to cover the scope and application of the terms of service set out by search service providers. This too is unnecessary because, via Schedule 8, the Bill already ensures transparency about the scope and application of the provisions that search services must make publicly available. I hope that gives the noble Lord some reassurance that the concerns he has raised are already covered. With that, I invite him to withdraw Amendment 16.

My Lords, I am very grateful to the Minister for that very detailed response, which I will have to read very carefully because it was quite complicated. That is the answer to my question. Terms of service will not be very easy to identify because to answer my questions he has had to pray in aid issues that Ofcom will necessarily have to assess—terms of services—to get at whether the companies are performing the duties that the Bill requires of them.

I will not go further on that. We know that there will be enough there to answer the main questions I had about this. I take the point about search being distinctively different in this area, although a tidy mind like mine likes to see all these things in one place and understand all the words. Every time I see “publicly available statement”, I do not know why but I think about people being hanged in public rather than a term of service or a contract.

The noble Lord, Lord Allan, made the point that nobody ever reads these terms of service. We generally agree with that, but if you are married to a lawyer, as I am, you read an awful lot more of these things than you perhaps feel are good for your diet. I cannot even go on holiday until I have proven to her that I have read every word of my insurance policy on what I will be shipped home with. It is a frightening thought that some people do that because they like doing it, and she does.

I will not take this much further. The jibe that I had at the beginning—that this does not quite fit with the rest of the Bill—is still there, but we will not get much change out of what we are doing. The important thing is that, even though it is a rather complicated route, it looks as though Ofcom will have, possibly retrospectively and with more transparency than actual powers, the ability to look at terms of service when they are not working.

What I miss is the ability to set a standard for terms of service that is broadly acceptable to people, which was exactly the point that the noble Lord made: they cannot be so complex that you will not read them but they have to be sufficient to achieve what they do. I am still lost about what you can use the triple shield for if you do not know whether the services will deliver what you know you do not want. I beg leave to withdraw the amendment.

Amendment 16 withdrawn.

Amendment 16A

Moved by

16A: Clause 8, page 7, line 23, after “19(2)” insert “and (8A)”

Member’s explanatory statement

This amendment inserts a signpost to the new duty in clause 19 about supplying records of risk assessments to OFCOM.

Amendment 16A agreed.

Clause 8, as amended, agreed.

Clause 9: Safety duties about illegal content

Amendments 16B and 16C

Moved by

16B: Clause 9, page 7, line 27, leave out “all”

Member’s explanatory statement

This is a technical amendment needed because the new duty to summarise illegal content risk assessments in the terms of service (see the amendment in the Minister’s name inserting new subsection (8A) below) is imposed only on providers of Category 1 services.

16C: Clause 9, page 7, line 27, at end insert “(as indicated by the headings).”

Member’s explanatory statement

This amendment provides clarification because the new duty to summarise illegal content risk assessments in the terms of service (see the amendment inserting new subsection (8A) below) is imposed only on providers of Category 1 services.

Amendments 16B and 16C agreed.

Amendment 17

Moved by

17: Clause 9, page 7, line 30, leave out “prevent individuals from” and insert “protect individuals from harms arising due to them”

Member’s explanatory statement

This amendment, along with the other amendment to Clause 9 in the name of Lord Moylan, adds a requirement to protect individuals from harm, rather than monitoring, prior restraint and/or denial of access. Further obligations to mitigate and manage harm, including to remove unlawful content that is signalled to the service provider, are unchanged by this amendment.

My Lords, this is a very large and wide-ranging group of amendments. Within it, I have a number of amendments that, on their own, span three separate subjects. I propose to address these one after the other in my opening remarks, but other subjects will be brought in as the debate continues and other noble Lords speak to their own amendments.

If I split the amendments that I am speaking to into three groups, the first is Amendments 17 and 18. These relate to Clause 9, on page 7, where safety duties about illegal content are set out. The first of those amendments addresses the obligation to prevent individuals encountering priority illegal content by means of the service.

Earlier this week in Committee, I asked the Minister whether the Government understood “prevent” and “protect”, both of which they use in the legislation, to have different weight. I did not expect my noble friend to give an answer at that point, but I know that he will have reflected on it. We need clarity about this at some point, because courts will be looking at, listening to and reading what the Government say at the Dispatch Box about the weight to be given to these words. To my mind, to prevent something happening requires active measures in advance that ensure as far as reasonably and humanly possible that it does not actually happen, but one could be talking about something more reactive to protect someone from something happening.

This distinction is of great importance to internet companies—I am not talking about the big platforms—which will be placed, as I say repeatedly, under very heavy burdens by the Bill. It is possible that they simply will not be able to discharge them and will have to go out of business.

Let us take Wikipedia, which was mentioned earlier in Committee. It operates in 300 languages but employs 700 moderators globally to check what is happening. If it is required by Clause 9 to

“prevent individuals from encountering priority illegal content by means of the service”,

it will have to scrutinise what is put up on this community-driven website as or before it appears. Quite clearly, something such as Welsh Wikipedia—there is Wikipedia in Welsh—simply would not get off the ground if it had to meet that standard, because the number of people who would have to be employed to do that would be far more than the service could sustain. However, if we had something closer to the wording I suggest in my amendment, where services have to take steps to “protect” people—so they could react to something and take it down when they become aware of it—it all becomes a great deal more tolerable.

Similarly, Amendment 18 addresses subsection (3) of the same clause, where there is a

“duty to operate a service using proportionate systems and processes … to … minimise the length of time”

for which content is present. How do you know whether you are minimising the length of time? How is that to be judged? What is the standard by which that is to be measured? Would it not be a great deal better and more achievable if the wording I propose, which is that you simply are under an obligation to take it down, were inserted? That is my first group of amendments. I put that to my noble friend and say that all these amendments are probing to some extent at this stage. I would like to hear how he thinks that this can actually be operated.

My second group is quite small, because it contains only Amendment 135. Here I am grateful to the charity JUSTICE for its help in drawing attention to this issue. This amendment deals with Schedule 7, on page 202, where the priority offences are set out. Paragraph 4 of the schedule says that a priority offence includes:

“An offence under any of the following provisions of the Public Order Act 1986”.

One of those is Section 5 of that Act, “Harassment, alarm or distress”. Here I make a very different point and return to territory I have been familiar with in the past. We debated this only yesterday in Grand Committee, although I personally was unable to be there: the whole territory of hate crimes, harmful and upsetting words, and how they are to be judged and dealt with. In this case, my amendment would remove Section 5 of the Public Order Act from the list of priority offences.

If society has enough problems tolerating the police going round and telling us when we have done or said harmful and hurtful things and upbraiding us for it, is it really possible to consider—without the widest form of censorship—that it is appropriate for internet platforms to judge us, shut us down and shut down our communications on the basis of their judgment of what we should be allowed to say? We already know that there is widespread suspicion that some internet platforms are too quick to close down, for example, gender critical speech. We seem to be giving them something close to a legislative mandate to be very trigger-happy when it comes to closing down speech by saying that it engages, or could engage, Section 5 of the Public Order Act. I will come to the question of how they judge it in my third group, in a moment—but the noble Lord might be able to help me.

Just to reinforce the point the noble Lord, Lord Moylan, made on that, I certainly had experience of where the police became the complainants. They would request, for example, that you take down an English Defence League event, claiming that it would be likely to cause a public order problem. I have no sympathy whatever with the English Defence League, but I am very concerned about the police saying “You must remove a political demonstration” to a platform and citing the legal grounds for doing that. The noble Lord is on to a very valid point to be concerned about that.

I am grateful to the noble Lord. I really wonder whether the Government realise what they are walking into here. On the one hand, yesterday the Grand Committee was debating the statutory instrument putting in place new statutory guidance for the police on how to enforce, much more sensitively than in the past, non-crime hate incidents. However, on the other hand, the next day in this Chamber we are putting an obligation on a set of mostly foreign private companies to act as a police force to go around bullying us and closing us down if we say something that engages Section 5 of the Public Order Act. I think this is something the Government are going to regret, and I would very much like to hear what my noble friend has to say about that.

Finally, I come to my third group of amendments: Amendments 274, 278, 279 and 283. They are all related and on one topic. These relate to the text of the Bill on page 145, in Clause 170. Here we are discussing what judgments providers have to make when they come to decide what material to take down. Inevitably, they will have to make judgments. That is one of the unfortunate things about this Bill. A great deal of what we do in our lives is going to have to be based on judgments made by private companies, many of which are based abroad but which we are trying to legislate for.

It makes a certain sense that the law should say what they should take account of in making those judgments. But the guidance—or rather, the mandate—given to those companies by Clause 170 is, again, very hair-trigger. Clause 170(5), which I am proposing we amend, states:

“In making such judgements, the approach to be followed is whether a provider has reasonable grounds to infer that content is … of the kind in question”.

I am suggesting that “reasonable grounds to infer” should be replaced with “sufficient evidence to infer”, so that they have to be able to produce some evidence that they are justified in taking content down. The test should be higher than simply having “reasonable grounds”, which may rest on a suspicion and little evidence at all. So one of those amendments relates to strengthening that bar so that they must have real evidence before they can take censorship action.

I add only two words to subsection (6), which talks about reasonable grounds for the inference—it defines what the reasonable grounds are—that

“exist in relation to content and an offence if, following the approach in subsection (2)”

and so on. I am saying “if and only if”—in other words, I make it clear that this is the only basis on which material can be censored using the provisions in this section, so as to limit it from going more widely. The third amendment in my group is essentially consequential to that.

We are all worried in this Committee about prospect of speech being censored in a way which infringes the freedom of speech rights that we fought so hard to establish and which are also embedded in Article 10 of the European Convention on Human Rights. We want to have a legal structure that does not empower providers to act as private sector censors ranging over what we do, except in circumstances where it is wholly justified and in the public interest. The language in the Bill is far too loose for this purpose. It does not give us the protection. It does not do what my noble friend said it would do when he spoke at Second Reading, which is to strike the right balance. These amendments in my third group—and indeed the one in my second group—are there to help strike the right balance. I beg to move.

My Lords, I will speak to Amendments 128, 130 and 132, as well as Amendments 143 to 153 in this grouping. They were tabled in the name of my right reverend colleague the Bishop of Derby, who is sorry that she cannot be here today.

The Church of England is the biggest provider of youth provision in our communities and educates around 1 million of our nation’s children. My colleague’s commitment to the principles behind these amendments also springs from her experience as vice chair of the Children’s Society. The amendments in this grouping are intended to strengthen legislation on online grooming for the purpose of child criminal exploitation, addressing existing gaps and ensuring that children are properly protected. They are also intended to make it easier for evidence of children being groomed online for criminal exploitation to be reported by online platforms to the police and the National Crime Agency.

Research from 2017 shows that one in four young people reported seeing illicit drugs advertised for sale on social media—a percentage that is likely to be considerably higher six years on. According to the Youth Endowment Fund in 2022, 20% of young people reported having seen online content promoting gang membership in the preceding 12 months, with 24% reporting content involving the carrying, use or promotion of weapons.

In relation to drugs, that later research noted that these platforms provide opportunities for dealers to build trust with potential customers, with young people reporting that they are more likely to see a groomer advertising drugs as a friend than as a dealer. This leaves young people vulnerable to exploitation, thereby reducing the scruples or trepidation they might feel about buying drugs in the first place. Meanwhile, it is also clear that social media is changing the operation of the county lines model. There is no longer the need to transport children from cities into the countryside to sell drugs, given that children who live in less populated areas can be groomed online as easily as in person. A range of digital platforms is therefore being used to target potential recruits among children and young people, with digital technologies also being deployed—for example, to monitor their whereabouts on a drugs run.

More research is being carried out by the Children’s Society, whose practitioners reported a notable increase in the number of perpetrators grooming children through social media and gaming sites during the first and second waves of the pandemic. Young people were being contacted with promotional material about lifestyles they could lead and the advantages of working within a gang, and were then asked to do jobs in exchange for money or status within this new group. It is true that some such offences could be prosecuted under the Modern Slavery Act 2015, but there remains a huge disparity between the scale of exploitation and the number of those being charged under the Act. Without a definition of child exploitation for criminal purposes, large numbers of children are being groomed online and paying the price for crimes committed by some of their most dangerous and unscrupulous elders.

It is vital that we protect our children from online content which facilitates that criminal exploitation, in the same way that we are looking to protect them from sexual exploitation. Platforms must be required to monitor for illegal content related to child criminal exploitation on their sites and to have mechanisms in place for users to flag it with those platforms so it can be removed. This can be achieved by including modern slavery and trafficking, of which child criminal exploitation is a form, into the scope of illegal content within the Bill, which is what these amendments seek to do. It is also vital that the law sets out clear expectations on platforms to report evidence of child criminal exploitation to the National Crime Agency in the same way as they are expected to report content involving child sexual exploitation and abuse to enable child victims to be identified and to receive support. Such evidence may enable action against the perpetrators without the need of a disclosure from child victims. I therefore fully support and endorse the amendments standing in the name of the right reverend Prelate.

My Lords, this is again a very helpful set of amendments. I want to share some experience that shows that legality tests are really hard. Often from the outside there is an assumption that it is easy to understand what is legal and illegal in terms of speech, but in practice that is very rarely the case. There is almost never a bright line, except in a small class of child sexual abuse material where it is always illegal and, as soon as you see the material, you know it is illegal and you can act on it. In pretty much every other case, you have to look at what is in front of you.

I will take a very specific example. Something we had to deal with was images of Abdullah Öcalan, the leader of the PKK in Turkey. If somebody shared a picture of Abdullah Öcalan, were they committing a very serious offence, which is the promotion of terrorism? Were they indicating support for the peace process that was taking place in Turkey? Were they showing that they support his socialist and feminist ideals? Were they supporting the YPG, a group in Syria to which we were sending arms, that venerates him? This is one example of many I could give where the content in front of you does not tell you very clearly whether or not the speech is illegal or speech that should be permitted. Indeed, we would take speech like that down and I would get complaints, including from Members of Parliament, saying, “Why have you removed that speech? I’m entitled to talk about Abdullah Öcalan”, and we would enter into an argument with them.

We would often ask lawyers in different countries whether they could tell us whether a speech was legal or illegal. The answer would come back as probably illegal, likely illegal, maybe illegal and, occasionally, definitely not illegal, but it was nearly always on the spectrum. The amendments we are proposing today are to try to understand where the Government intend people to draw that line when they get that advice. Let us assume the company wants to do the right thing and follow the instructions of the Bill and remove illegal content. At what level do they say it has met the test sufficiently, given that in the vast majority of cases, apart from the small class of illegal content, they are going to be given only a likelihood or a probability? As the noble Lord, Lord Moylan, pointed out, we have to try to insert this notion of sufficient evidence with Amendments 273, 275, 277, 280 and 281 in the names of my noble friend Lord Clement-Jones and the noble Viscount, Lord Colville, who is unable to be in his place today. I think the noble Baroness, Lady Kidron, may also have signed them. We are trying to flesh out the point at which that illegality standard should kick in.

Just to understand again how this often works when the law gets involved, I say that there is a law in Germany; the short version is NetzDG. If there are any German speakers who can pronounce the compound noun that is its full title, there will be a prize. It is a long compound word that means “network enforcement Act”. It has been in place for a few years and it tells companies to do something similar—to remove content that is illegal in Germany. There would be cases where we would get a report from somebody saying, “This is illegal”, and we would take action; then it went into the German system and three months later we would finally get told whether it was actually illegal in a 12-page judgment that a German court had figured out. In the meantime, all we could do was work on our best guess while that process was going on. I think we need to be very clear that illegality is hard.

Cross-jurisdictional issues present us with another set of challenges. If both the speaker and the audience are in the United Kingdom, it is fairly clear. But in many cases, when we are talking about online platforms, one or other, or even both the speaker and the audience, may be outside the United Kingdom. Again, when does the speech become illegal? It may be entirely legal speech between two people in the United States. I think—and I would appreciate clarification from the Minister—that the working assumption is that if the speech was reported by someone not in the United State but in the UK, the platform would be required to restrict access to it from the UK, even though the speech is entirely legal in the jurisdiction in which it took place. Because the person in the UK encountered it, there would be a duty to restrict it. Again, it has been clarified that there is certainly not a duty to take the speech down, because it is entirely legal speech outside the UK. These cross-jurisdictional issues are interesting; I hope the Minister can clarify that.

The amendments also try to think about how this would work in practice. Amendment 287 talks about how guidance should be drawn up in consultation with UK lawyers. That is to avoid a situation where platforms are guessing too much at what UK lawyers want; they should at least have sought UK legal advice. That advice will then be fed into the guidance given to their human reviewers and their algorithms. That is the way, in practice, in which people will carry out the review. There is a really interesting practical question—which, again, comes up under NetzDG—about the extent to which platforms should be investing in legal review of content that is clearly against their terms of service.

There will be two kinds of platform. There will be some platforms that see themselves as champions of freedom of expression and say they will only remove stuff that is illegal in the UK, and everything else can stay up. I think that is a minority of platforms—they tend to be on the fringes. As soon as a platform gets a mainstream audience, it has to go further. Most platforms will have terms of service that go way beyond UK law. In that case, they will be removing the hate speech, and they will be confident that they will remove UK-illegal hate speech within that. They will remove the terrorist content. They will be confident and will not need to do a second test of the legality in order to be able to remove that content. There is a practical question about the extent to which platforms should be required to do a second test if something is already illegal under their terms.

There will be, broadly speaking again, four buckets of content. There will be content that is clearly against a platform’s terms, which it will want to get rid of immediately. It will not want to test it again for legality; it will just get rid of it.

There will be a second bucket of content that is not apparently against a platform’s terms but clearly illegal in the UK. That is a very small subset of content: in Germany, that is Holocaust denial content; in the United Kingdom, this Parliament has looked at Holocaust denial and chosen not to criminalise it, so that will not be there, but an equivalent for us would be migration advice. Migration advice will not be against the terms of service of most platforms, but in the Government’s intention, the Illegal Migration Bill is to make it illegal and require it to be removed, and the consequent effect will be that it will have to be removed under the terms of this Bill. So there will be that small set of content that is illegal in the UK but not against terms of service.

There will be a third bucket of content that is not apparently against the terms or the law, and that actually accounts for most of the complaints that a platform gets. I will choose my language delicately: complaint systems are easy, and people complain to make a point. They use complaint systems such as dislike buttons. The reality is that one of the most common sets of complaints you get is when there is a football match and the two opposing teams report the content on each other’s pages as illegal. They will do that every time, and you get used to it, and that is why you learn to discount mass-volume complaints. But again, we should be clear that there are a great many complaints that are merely vexatious.

The final bucket is of content that is unclear and legal review will be needed. Our amendment is intended to deal with those. A platform will go out and get advice. It is trying to understand at what point something like migration advice tips over into the illegal as opposed to being advice about going on holiday, and it is trying to understand that based on what it can immediately see. Once it has sought that advice, it will feed that back into the guidance to reviewers and the algorithms to try and remove content more effectively and be compliant with the Bill as a whole and not get into trouble with Ofcom.

Some areas are harder than others. The noble Lord, Lord Moylan, already highlighted one: public order offences, which are extremely hard. If somebody says something offensive or holds an offensive political view—I suspect the noble Baroness, Lady Fox, may have something to say on this—people may well make contact and claim that it is in breach of public order law. On the face of it, they may have a reasonably arguable case but again, as a platform, you are left to make a decision.

There is a really interesting potential role for Ofcom here. One thing that is frustrating if you work at a platform is that you will often get stuck and when you go out and look for advice, you find it is hard to get it. When I ran a working group with some French lawyers, including quite senior judges, they came into the working group saying, “This is all straightforward—you’re just not removing the illegal stuff”. So we gave them real cases and it was interesting to see how half of the lawyers in the room would be on one side, saying “It must come down—it’s against French law” while the other half was saying, “How could you possibly take this down in France?”, because it was protected speech. It is really difficult to get that judgment but, interestingly, an unintended consequence of the Bill may be that Ofcom will ultimately get stuck in that position.

The Bill is not about Ofcom making rulings on individual items of content but if—as in the example I shared with the noble Lord, Lord Moylan, earlier—the police have said to a platform, “You must remove this demonstration. It is illegal”, and the platform said, “No, we judge it not to be illegal”, where are the police going to go? They will go to Ofcom and say, “Look, this platform is breaching the law”, so Ofcom is going to get pulled into that kind of decision-making. I do not envy it that but, again, we need to plan for that scenario because people who complain about illegality will go wherever they think they can get a hearing, and Ofcom will be one of those entities.

A huge amount on this illegal content area still needs to be teased out. I ask the Minister to respond specifically to the points I have raised around whose jurisdiction it is. If the speaker is speaking legally, because they are in a country outside the United Kingdom, what is the Government’s expectation on platforms in those circumstances? Will he look at the issue of the tests and where on this spectrum, from probably illegal through to likely to be illegal and may be illegal, the Government expect platforms to draw the line? If platforms have removed the bad content, will he consider carefully to what extent the Government think that the platforms should have to go through the process of investing time and energy to work out whether they removed it for illegality or for a terms of service breach? That is interesting but if our focus is on safety, frankly, it is wasted effort. We need to question how far we expect the platforms to do that.

My Lords, before speaking to my Amendment 137, I want to put a marker down to say that I strongly support Amendment 135 in the name of my noble friend Lord Moylan. I will not repeat anything that he said but I agree with absolutely every word.

Amendment 137 is in my name and that of my noble and learned friend Lord Garnier and the noble Lord, Lord Moore of Etchingham. This amendment is one of five which I have tabled with the purpose of meeting a core purpose of the Bill. In the words of my noble friend the Minister in response to Amendment 1, it is

“to protect users of all ages from being exposed to illegal content”—[Official Report, 19/4/23; col. 724.]

—in short, to ensure that what is illegal offline is illegal online.

If accepted, this small group of amendments would, I strongly believe, make a really important difference to millions of people’s lives—people who are not necessarily listed in Clause 12. I therefore ask the Committee to allow me to briefly demonstrate the need for these amendments through the prism of millions of people and their families working and living in rural areas. They are often quite isolated and working alone in remote communities, and are increasingly at risk of or are already suffering awful online abuse and harassment. This abuse often goes way beyond suffering; it destroys businesses and a way of life.

I find it extraordinary that the Bill seems to be absent of anything to do with livelihoods. It is all about focusing on feelings, which of course are important—and the most important focus is children—but people’s businesses and livelihoods are being destroyed through abuse online.

Research carried out by the Countryside Alliance has revealed a deeply disturbing trend online that appears to be disproportionately affecting people who live in rural areas and who are involved in rural pursuits. Beyond direct abuse, a far more insidious tactic that activists have adopted involves targeting businesses involved in activities of which they disapprove, such as livestock farming or hosting shoots. They post fake reviews on platforms including Tripadvisor and Google Maps, and their aim is to damage the victim, their business and their reputation by, to put it colloquially, trashing their business and thereby putting off potential customers. This is what some call trolling.

Let me be clear that I absolutely defend, to my core, the right to freedom of expression and speech, and indeed the right to offend. Just upsetting someone is way below the bar for the Bill, or any legislation. I am deeply concerned about the hate crime—or non-crime—issue we debated yesterday; in fact, I put off reading the debate because I so disagree with this nonsense from the College of Policing.

Writing a negative review directly based on a negative experience is entirely acceptable in my book, albeit unpleasant for the business targeted. My amendments seek to address something far more heinous and wrong, which, to date, can only be addressed as libel and, therefore, through the civil courts. Colleagues in both your Lordships’ House and in another place shared with me tremendously upsetting examples from their constituents and in their neighbourhoods of how anonymous activists are ruining the lives of hard-working people who love this country and are going the extra mile to defend our culture, historic ways of life and freedoms.

Fortunately, through the Bill, the Government are taking an important step by introducing a criminal offence of false communications. With the leave of the Committee, I will briefly cite and explain the other amendments in order to make sense of Amendment 137. One of the challenges of the offence of false communications is the need to recognise that so much of the harm that underpins the whole reason why the Bill is necessary is the consequence of allowing anonymity. It is so easy to destroy and debilitate others by remaining anonymous and using false communications. Why be anonymous if you have any spine at all to stand up for what you believe? It is not possible offline—when writing a letter to a newspaper, for example—so why is it acceptable online? The usual tech business excuse of protecting individuals in rogue states is no longer acceptable, given the level of harm that anonymity causes here at home.

Therefore, my Amendment 106 seeks to address the appalling effect of harm, of whatever nature, arising from false or threatening communications committed by unverified or anonymous users—this is what we refer to as trolling. Amendments 266 and 267, in my name and those of my noble and learned friend Lord Garnier and my noble friend Lord Leicester, would widen the scope of this new and welcome offence of false communications to include financial harm, and harm to the subject of the false message arising from its communication to third parties.

The Bill will have failed unless we act beyond feelings and harm to the person and include loss of livelihood. As I said, I am amazed that it is not front and centre of the Bill after safety for our children. Amendment 268, also supported by my noble and learned friend, would bring within the scope of the communications offences the instigation of such offences by others—for example, Twitter storms, which can involve inciting others to make threats without doing so directly. Currently, we are unsure whether encouraging others to spread false information—for example, by posting fake reviews of businesses for ideologically motivated reasons—would become an offence under the Bill. We believe that it should, and my Amendment 268 would address this issue.

I turn briefly to the specifics of my Amendment 137. Schedule 7 lists a set of “priority offences” that social media platforms must act to prevent, and they must remove messages giving rise to certain offences. However, the list does not include the new communications offences created elsewhere in Part 10. We believe that this is a glaring anomaly. If there is a reason why the new communications offences are not listed, it is important that we understand why. I hope that my noble friend the Minister can explain.

The practical effect of Amendment 137 would be to include the communications offences introduced in the Bill and communications giving rise to them within the definition of “relevant offence” and “priority illegal content” for the purposes of Clause 53(4) and (7) and otherwise.

I ask the Committee to have a level of imagination here because I have been asked to read the speech of the noble Viscount, Lord Colville—

I do not know who advised the noble Baroness—and forgive me for getting up and getting all former Leader on her—but this is a practice that we seem to have adopted in the last couple of years and that I find very odd. It is perfectly proper for the noble Baroness to deploy the noble Viscount’s arguments, but to read his speech is completely in contravention of our guidance.

I beg the pardon of the Committee. I asked about it and was misinformed; I will do as the noble Baroness says.

The noble Viscount, Lord Colville, is unable to be with us. He put his name to Amendments 273, 275, 277 and 280. His concern is that the Bill sets the threshold for illegality too low and that in spite of the direction provided by Clause 170, the standards for determining illegality are too vague.

I will make a couple of points on that thought. Clause 170(6) directs that a provider must have

“reasonable grounds to infer that all elements necessary for the commission of the offence, including mental elements, are present or satisfied”,

but that does not mean that the platform has to be certain that the content is illegal before it takes it down. This is concerning when you take it in combination with what or who will make judgments on illegality.

If a human moderator makes the decision, it will depend on the resources and time available to them as to how much information they gather in order to make that judgment. Unlike in a court case, when a wide range of information and context can be gathered, when it comes to decisions about content online, these resources are very rarely available to human moderators, who have a vast amount of content to get through.

If an automated system makes the judgment, it is very well established that algorithms are not good at context—the Communications and Digital Committee took evidence on this repeatedly when I was on it. AI simply uses the information available in the content itself to make a decision, which can lead to significant missteps. Clause 170(3) provides the requirement for the decision-makers to judge whether there is a defence for the content. In the context of algorithms, it is very unclear how they will come to such a judgment from the content itself.

I understand that these are probing amendments, but I think the concern is that the vagueness of the definition will lead to too much content being taken down. This concern was supported by Parliament’s Joint Committee on Human Rights, which wrote to the former Culture Secretary, Nadine Dorries, on that matter. I apologise again.

My Lords, I support the amendments in this group that probe how removing illegal material is understood and will be used under the Bill. The noble Lord, Lord Moylan, explained a lot of my concerns, as indeed did the noble Viscount, Lord Colville, via his avatar. We have heard a range of very interesting contributions that need to be taken seriously by the Government. I have put my name to a number of amendments.

The identification of illegal material might be clear and obvious in some cases—even many cases. It sounds so black and white: “Don’t publish illegal material”. But defining communications of this nature can be highly complex, so much so that it is traditionally reserved for law enforcement bodies and the judicial system. We have already heard from the noble Lord, Lord Moylan, that, despite Home Secretaries, this House, regulations and all sorts of laws having indicated that non-crime hate incidents, for example, should not be pursued by the police, they continue to pursue them as though they are criminal acts. That is exactly the kind of issue we have.

I noted earlier that the noble Lord, Lord Bethell, made a passionate intervention about, of all things, Andrew Tate and his illegality in relation to this Bill. That prompted me to think a number of things. Andrew Tate is an influencer who I despise, as I do the kind of things he says. But, as far as I know, the criminal allegations he faces are not yet resolved, so he has to be seen as innocent until proven guilty. Most of what he has online that is egregious might well be in bad taste, as people say—I would say that it is usually misogynist—but it is not against the law. If we get to a situation where that is described as illegality, that is the kind of thing that I worry about. As we have heard from other noble Lords, removing so-called illegal content for the purpose of complying with this regulatory system will mean facing such dilemmas.

In talking about individuals and investigations, the noble Baroness reminded me of one class of content where we do have clarity, and that is contempt of court. That is a frequent request. We know that it is illegal in that case because a judge writes to the company and says, “You must not allow this to be said because it is in contempt of court”, but that really is the exception. In most other cases, someone is saying, “I think it is illegal”. In live proceedings, in most cases it is absolutely clear because a judge has told you.

That is very helpful.

I am concerned that removing so-called illegal content for the purpose of complying with the regulatory system covers not only that which reaches conviction in a criminal court but possibly anything that a platform determines could be illegal, and therefore it undermines our own legal system. As I have said, that marks a significant departure from the rule of law. It seems that the state is asking or mandating private companies to make determinations about what constitutes illegality.

The obligations on a platform to determine what constitutes illegality could obviously become a real problem, particularly in relation to limitations on free expression. As we have already heard, the Public Order Act 1986 criminalises, for example, those who stir up hatred through the use of words, behaviour or written material. That is contentious in the law offline. By “contentious”, I mean that it is a matter of difficulty that requires the full rigour of the criminal justice system, understanding the whole history of established case law. That is all necessary to make a conviction under that law for offences of this nature.

Now we appear to be saying that, without any of that, social media companies should make the decision, which is a nerve-racking situation to be in. We have already heard the slippery phrase “reasonable grounds to infer”. If that was the basis on which you were sent to prison—if they did not have to prove that you were guilty but they had reasonable grounds to infer that you might be, without any evidence—I would be worried, yet reasonable grounds to infer that the content could be illegal is the basis on which we are asking for those decisions to be made. That is significantly below the ordinary burden of proof required to determine that an illegal act has been committed. Under this definition, I fear that platforms will be forced to overremove and censor what ultimately will be entirely lawful speech.

Can the Minister consider what competency social media companies have to determine what is lawful? We have heard some of the dilemmas from somebody who was in that position—let alone the international complications, as was indicated. Will all these big tech companies have to employ lots of ex-policemen and criminal lawyers? How will it work? It seems to me that there is a real lack of qualifications in that sphere— that is not a criticism, because those people decided to work in big tech, not in criminal law, and yet we are asking them to pursue this. That is a concern.

I will also make reference to what I think are the controversies around government Amendments 136A and 136B to indicate the difficulties of these provisions. They concern illegal activity—such as “assisting unlawful immigration”, illegal entry, human trafficking and similar offences—but I am unsure as to how this would operate. While it is the case that certain entrances to the UK are illegal, I suddenly envisage a situation where a perfectly legitimate political debate—for example, about the small boats controversy—would be taken down, and that people advocating for a position against the Government’s new Illegal Migration Bill could be accused of supporting illegality. What exactly will be made illegal in those amendments to the Online Safety Bill?

The noble Baroness, Lady Buscombe, made a fascinating speech about an interesting group of amendments. Because of the way the amendments are grouped, I feel that we have moved to a completely different debate, so I will not go into any detail on this subject. Anonymous trolling, Twitter storms and spreading false information are incredibly unpleasant. I am often the recipient of them—at least once a week—so I know personally that you feel frustrated that people tell lies and your reputation is sullied. However, I do not think that these amendments offer the basis on which that activity should be censored, and I will definitely argue against removing anonymity clauses—but that will be in another group. It is a real problem, but I do not think that the solution is contained in these amendments.

My Lords, my contribution will be less officious than my intervention earlier in this group. In the last couple of years since I returned to the House—as I describe it—having spent time at the Charity Commission, I have noticed a new practice emerging of noble Lords reading out other people’s speeches. Every time I had seen it happen before, I had not said anything, but today I thought, “I can’t sit here and not say anything again”. I apologise for my intervention.

I am grateful to my noble friend Lord Moylan for bringing forward his amendments and for introducing them in the incredibly clear way he did; they cover some very complex and diverse issues. I know that there are other amendments in the group which might be described as similar to his.

There are a couple of things I want to highlight. One interesting thing about the debate on this group is the absence of some of our legal friends—I apologise to my noble friend Lady Buscombe, who is of course a very distinguished lawyer. The point I am making is that we are so often enriched by a lot of legal advice and contributions on some of the more challenging legal issues that we grapple with, but we do not have that today, and this is a very difficult legal issue.

It is worth highlighting again, as has been touched on a little in some of the contributions, the concern, as I understand it, with how the Bill is drafted in relation to illegal content and the potential chilling effect of these clauses on social media platforms. As has already been said, there is a concern that it might lead them to take a safety-first approach in order to avoid breaking the law and incurring the sanctions and fines that come with the Bill, which Ofcom will have the power to apply. That is the point we are concerned with here. It is the way in which this is laid out, and people who are much better equipped than I am have already explained the difference between evidence versus reasonable grounds to infer.

What the noble Lord, Lord Allan, hit on in his contribution is also worth taking into account, and that is the role of Ofcom in this situation. One of the things I fear, as we move into an implementation phase and the consequences of the Bill start to impact on the social media firms, is the potential for the regulator to be weaponised in a battle on the cultural issues that people are becoming increasingly exercised about. I do not have an answer to this, but I think it is important to understand the danger of where we might get to in the expectations of the regulator if we create a situation where the social media platforms are acting in a way that means people are looking for recourse or a place to generate further an argument and a battle that will not be helpful at all.

I am not entirely sure, given my lack of legal expertise —this is why I would have been very grateful for some legal expertise on this group—whether what my noble friend is proposing in his amendments is the solution, but I think we need to be very clear that this is a genuine problem. I am not sure, as things stand in the Bill, that we should be comfortable that it is not going to create problems. We need to find a way to be satisfied that this has been dealt with properly.

It is a great honour to follow my noble friend. I completely agree with her that this is a powerful discussion and there are big problems in this area. I am grateful also to my noble friend Lord Moylan for raising this in the first place. It has been a very productive discussion.

I approach the matter from a slightly different angle. I will not talk about the fringe cases—the ones where there is ambiguity, difficulty of interpretation, or responsibility or regulatory override, all of which are very important issues. The bit I am concerned about is where primary priority content that clearly demonstrates some kind of priority offence is not followed up by the authorities at all.

The noble Lord, Lord Allan, referred to this point, although he did slightly glide over it, as though implying, if I understood him correctly, that this was not an area of concern because, if a crime had clearly been committed, it would be followed up on. My fear and anxiety is that the history of the internet over the last 25 years shows that crimes—overt and clear crimes that are there for us to see—are very often not followed up by the authorities. This is another egregious example of where the digital world is somehow exceptionalised and does not have real-world rules applied to it.

The noble Baroness, Lady Fox, quite reasonably asked me about Andrew Tate. That matter is sub judice; the noble Lord, Lord Allan, referred to it and I do not want to drag the conversation into dangerous legal territory. However, she makes the good point that we sometimes see, particularly in the online abuse of women, offences that are quite clearly crimes; they are crimes of rape, violent abuse and child abuse. It would not take any of us long to find videos that showed clear examples of crime, but very often they are not followed up with the energy and determination that they could or should be, because things on the internet somehow do not seem to touch the authorities in the way they should do.

His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services—

I want to clarify one point. I have had a slightly different experience, which is that for many people—women, at least—whom I have talked to recently, there is an over-enthusiasm and an over-zealous attitude to policing the speech of particular women and, as we have already heard, gender-critical women. It is often under the auspices of hate speech and there is all sorts of discussion about whether the police are spending too long trawling through social media. By contrast, if you want to get a policeman or policewoman involved in a physical crime in your area, you cannot get them to come out. So I am not entirely convinced. I think policing online speech at least is taking up far too much of the authorities’ time, not too little time, and distracting them from solving real social and criminal activity.

I defer to the noble Baroness, Lady Fox, on speech crime. That is not the area of my expertise, and it is not the purpose of my points. My points were to do with the kinds of crime that affect children in particular. His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services is very specific about that point. It says that “unacceptable delays are commonplace” and it gives a very large number of case studies. I will not go through them now because it is Thursday afternoon, but I think noble Lords can probably imagine the kinds of things we are talking about. They include years of delay, cases not taken seriously or overlooked, evidence lost, and so forth. The report found that too often children were put at risk because of this, and offenders were allowed to escape justice, and it gave 17 recommendations for how the police force should adapt in order to meet this challenge.

So my questions to the Minister are these. When we talk about things such as age verification for hardcore porn, we are quite often told that we do not need to worry about some of this because it is covered by illegal content provisions, and we should just leave it to the police to sort out. His Majesty’s Inspectorate gives clear evidence—this is a recent report from last month—that this is simply not happening in the way it should be. I therefore wondered what, if anything, is in the Bill to try to close down this particular gap. That would be very helpful indeed.

If it is really not for the purposes of this Bill at all—if this is actually to do with other laws and procedures, other departments and the way in which the resources for the police are allocated, as the noble Baroness, Lady Fox, alluded to—what can the Government do outside the boundaries of this legislation to mobilise the police and the prosecution services to address what I might term “digital crimes”: that is, crimes that would be followed up with energy if they occurred in the real world but, because they are in the digital world, are sometimes overlooked or forgotten?

My Lords, I would like to mention one issue that I forgot to mention, and I think it would be more efficient to pose the question now to the Minister rather than interject when he is speaking.

On the Government’s Amendments 136A, 136B and 136C on the immigration offences, the point I want to make is that online services can be literal life-savers for people who are engaged in very dangerous journeys, including journeys across the Channel. I hope the Minister will be clear that the intention here is to require platforms to deal only with content, for example, from criminals who are offering trafficking services, and that there is no intention to require platforms somehow to withdraw services from the victims of those traffickers when they are using those services in the interest of saving their own lives or seeking advice that is essential to preserving their own safety.

That would create—as I know he can imagine—real ethical and moral dilemmas, and we should not be giving any signal that we intend to require platforms to withdraw services from people who are in desperate need of help, whatever the circumstances.

My Lords, we seem to have done it again—a very long list of amendments in a rather ill-conceived group has generated a very interesting discussion. We are getting quite good at this, exchanging views across the table, across the Committee, even within the Benches—Members who perhaps have not often talked together are sharing ideas and thoughts, and that is a wonderful feeling.

I want to start with an apology. I think I may be the person who got the noble Baroness, Lady Kidron, shopped by the former leader—once a leader, always a leader. What I thought I was being asked was whether the Committee would be interested in hearing the views of the noble Viscount who could not be present, and I was very keen, because when he does speak it is from a point of view that we do not often hear. I did not know that it was a transgression of the rules—but of course it is not, really, because we got round it. Nevertheless, I apologise for anything that might have upset the noble Baroness’s blood pressure—it did not stop her making a very good contribution later.

We have covered so much ground that I do not want to try and summarise it in one piece, because you cannot do that. The problem with the group as it stands is that the right reverend Prelate the Bishop of Derby and myself must have some secret connection, because we managed to put down almost the same amendments. They were on issues that then got overtaken by the Minister, who finally got round to—I mean, who put down a nice series of amendments which exactly covered the points we made, so we can lose all those. But this did not stop the right reverend Prelate the Bishop of Guildford making some very good additional points which I think we all benefited from.

I welcome back the noble Baroness, Lady Buscombe, after her illness; she gave us a glimpse of what is to come from her and her colleagues, but I will leave the particular issue that she raised for the Minister to respond to. It raises an issue that I am not competent on, but it is a very important one—we need to get the right balance between what is causing the alarm and difficulty outside in relation to what is happening on the internet, and I think we all agree with her that we should not put any barrier in the way of dealing with that.

Indeed, that was the theme of a number of the points that have been raised on the question of what is or can constitute illegal content, and how we judge it. It is useful to hear again from the master about how you do it in practice. I cannot imagine being in a room of French lawyers and experts and retaining my sanity, let alone making decisions that affect the ability of people to carry on, but the noble Lord did it; he is still here and lives to tell the tale—bearded or otherwise.

The later amendments, particularly from the noble Lord, Lord Clement-Jones, are taking us round in a circle towards the process by which Ofcom will exercise the powers that it is going to get in this area. These are probably worth another debate on their own, and maybe it will come up in a different form, because—I think the noble Baroness, Lady Stowell, made this point as well—there is a problem in having an independent regulator that is also the go-to function for getting advice on how others have to make decisions that are theirs to rule on at the end if they go wrong. That is a complicated way of saying that we may be overloading Ofcom if we also expect it to provide a reservoir of advice on how you deal with the issues that the Bill puts firmly on the companies—I agree that this is a problem that we do not really have an answer to.

My amendments were largely overtaken by the Government’s amendments, but the main one I want to talk about was Amendment 272. I am sorry that the noble Baroness, Lady Morgan, is not here, because her expertise is in an area that I want to talk about, which is fraud—cyber fraud in particular—and how that is going to be brought into the Bill. The issue, which I think has been raised by Which?, but a number of other people have also written to us about it, is that the Bill in Clauses 170 and 171 is trying to establish how a platform should identify illegal content in relation to fraud—but it is quite prescriptive. In particular, it goes into some detail which I will leave for the Minister to respond to, but uniquely it sets out a specific way for gathering information to determine whether content is illegal in this area, although it may have applicability in other areas.

One of the points that have to be taken into account is whether the platform is using human moderators, automated systems or a combination of the two. I am not quite sure why that is there in the Bill; that is really the basis for the tabling of our amendments. Clearly, one would hope that the end result is whether or not illegality has taken place, not how that information has been gathered. If one must make concessions to the process of law because a judgment is made that, because it is automated, it is in some way not as valid as if it had been done by a human moderator, there seems to be a whole world there that we should not be going into. I certainly hope that that is not going to be the case if we are talking about illegality concerning children or other vulnerable people, but that is how the Bill reads at present; I wonder whether the Minister can comment on that.

There is a risk of consumers being harmed here. The figures on fraud in the United Kingdom are extraordinary; the fact that it is not the top priority for everybody, let alone the Government, is extraordinary. It is something like the equivalent of consumers being scammed at the rate of around £7.5 billion per year. A number of awful types of scamming have emerged only because of the internet and social media. They create huge problems of anxiety and emotional distress, with lots of medical care and other things tied in if you want to work out the total bill. So we have a real problem here that we need to settle. It is great that it is in the Bill, but it would be a pity if the movement towards trying to resolve it is in any way infringed on by there being imperfect instructions in the Bill. I wonder whether the Minister would be prepared to respond to that; I would be happy to discuss it with him later, if that is possible.

As a whole, this is an interesting question as we move away from what a crime is towards how people judge how to deal with what they think is a crime but may not be. The noble Lord, Lord Allan, commented on how to do it in practice but one hopes that any initial problems will be overcome as we move forward and people become more experienced with this.

When the Joint Committee considered this issue, we spent a long time talking about why we were concerned about having certainty on the legal prescription in the Bill; that is why we were very much against the idea of “legal but harmful” because it seemed too subjective and too subject to difficulties. Out of that came another thought, which answers the point made by the noble Baroness, Lady Stowell: so much of this is about fine judgments on certain things that are there in stone and that you can work to but you then have to interpret them.

There is a role for Parliament here, I think; we will come on to this in later amendments but, if there is a debate to be had on this, let us not forget the points that have been made here today. If we are going to think again about Ofcom’s activity in practice, that is the sort of thing that either a Joint Committee or Select Committees of the two Houses could easily take on board as an issue that needs to be reflected on, with advice given to Parliament about how it might be taken forward. This might be the answer in the medium term.

In the short term, let us work to the Bill and make sure that it works. Let us learn from the experience but let us then take time out to reflect on it; that would be my recommendation but, obviously, that will be subject to the situation after we finish the Bill. I look forward to hearing the Minister’s response.

My Lords, as well as throwing up some interesting questions of law, this debate has provoked some interesting tongue-twisters. The noble Lord, Lord Allan of Hallam, offered a prize to the first person to pronounce the Netzwerkdurchsetzungsgesetz; I shall claim my prize in our debate on a later group when inviting him to withdraw his amendment.

I thank the noble Lord.

I was pleased to hear about Wicipedia Cymraeg—there being no “k” in Welsh. As the noble Lord, Lord Stevenson, said, there has been a very good conversational discussion in this debate, as befits Committee and a self-regulating House. My noble friend Lady Stowell is right to point out matters of procedure, although we were grateful to know why the noble Viscount, Lord Colville, supports the amendments in question.

My noble friend Lord Moylan’s first group within a group—Amendments 17 and 18—alters the duties in Clause 9 of the Bill. These amendments would weaken the illegal content duties by removing any obligation on services to take upstream measures to remove illegal content, including child sexual abuse material. They would therefore seriously undermine the Bill’s focus on proactive risk management. Similarly, Amendments 272 to 283 seek to alter how services should judge what is illegal. I understand that noble Lords are concerned, rightly, about the over-removal of content.

The amendments tabled by the noble Lord, Lord Clement-Jones, would require providers to have sufficient evidence that content is illegal before taking action against it, replacing the current test of “reasonable grounds to infer”. Sufficient evidence is a subjective measure. We have discussed the difficulties for those who must make these decisions and we think that this formulation would set an unclear threshold for providers to determine how they should judge illegality, which could result in the under-removal of illegal content, putting users at risk, or the over-removal of it, with adverse consequences for freedom of expression.

The amendments tabled by my noble friend Lord Moylan would narrow the test to require the removal only of content which, based on all reasonably available contextual evidence, is manifestly illegal, and we think that that threshold is too high. Context and analysis can give a provider good reasons to infer that content is illegal even though the illegality is not immediately obvious. This is the case with, for example, some terrorist content which is illegal only if shared with terrorist purposes in mind, and intimate image abuse, where additional information or context is needed to know whether content has been posted against the subject’s wishes.

Amendment 276 would remove the detail in Clause 170 that specifies the point at which providers must treat content as illegal or fraudulent. That would enable providers to interpret their safety duties in broader ways. Rather than having greater discretion, Ofcom would be given less certainty about whether it could successfully take enforcement action. I take the point raised by noble Lords about the challenges of how platforms will identify illegal content, and I agree with my noble friend Lady Stowell that the contributions of noble and learned Lords would be helpful in these debates as well. However, Clause 170 sets out how companies should determine whether or not content is illegal or an advertisement is fraudulent. I will say a little more about the context behind that, as the noble Lord, Lord Allan, may have a question.

The Bill recognises that it will often be difficult for providers to make judgments about content without considering the context. Clause 170 therefore clarifies that providers must ascertain whether, on the basis of on all reasonably available information, there are reasonable grounds to infer that all the relevant elements of the offence—including the mental elements—are present and that no defence is available. The amount of information that would be reasonably available to a particular service provider will depend on the size and capacity of the provider, among other factors.

Companies will need to ensure that they have effective systems to enable them to check the broader context relating to content when deciding whether or not to remove it. This will provide greater certainty about the standard to be applied by providers when assessing content, including judgments about whether or not content is illegal. We think that protects against over-removal by making it clear that platforms are not required to remove content merely on the suspicion of it being illegal. Beyond that, the framework also contains provisions about how companies’ systems and processes should approach questions of mental states and defences when considering whether or not content is an offence in the scope of the Bill.

I am struggling a little to understand why the Minister thinks that sufficient evidence is subjective, and therefore, I assume, reasonable grounds to infer is objective. Certainly, in my lexicon, evidence is more objective than inference, which is more subjective. I was reacting to that word. I am not sure that he has fully made the case as to why his wording is better.

I take the noble Lord’s point and my noble friend’s further contribution. I will see whether I can give a clearer and more succinct description in writing to flesh that out, but that it is the reason that we have alighted on the words that we have.

The noble Lord, Lord Allan, also asked about jurisdiction. If an offence has been committed in the UK and viewed by a UK user, it can be treated as illegal content. That is set out in Clause 53(11), which says:

“For the purposes of determining whether content amounts to an offence, no account is to be taken of whether or not anything done in relation to the content takes place in any part of the United Kingdom”.

I hope that that bit, at least, is clearly set out to the noble Lord’s satisfaction. It looks like it may not be.

Again, I think that that is clear. I understood from the Bill that, if an American says something that would be illegal were they to be in the United Kingdom, we would still want to exclude that content. But that still leaves it open, and I just ask the question again, for confirmation. If all of the activities are outside the United Kingdom—Americans talking to each other, as it were—and a British person objects, at what point would the platform be required to restrict the content of the Americans talking to each other? Is it pre-emptively or only as and when somebody in the United Kingdom objects to it? We should flesh out that kind of practical detail before this becomes law.

If it has been committed in the UK and is viewed by a UK user, it can be treated as illegal. I will follow up on the noble Lord’s further points ahead of the next stage.

Amendment 272 explicitly provides that relevant information that is reasonably available to a provider includes information submitted by users in complaints. Providers will already need to do this when making judgments about content, as it will be both relevant and reasonably available.

My noble friend Lord Moylan returned to the question that arose on day 2 in Committee, querying the distinction between “protect” and “prevent”, and suggesting that a duty to protect would or could lead to the excessive removal of content. To be clear, the duty requires platforms to put in place proportionate systems and processes designed to prevent users encountering content. I draw my noble friend’s attention to the focus on systems and processes in that. This requires platforms to design their services to achieve the outcome of preventing users encountering such content. That could include upstream design measures, as well as content identification measures, once content appears on a service. By contrast, a duty to protect is a less stringent duty and would undermine the proactive nature of the illegal content duties for priority offences.

Before he moves on, is my noble friend going to give any advice to, for example, Welsh Wikipedia, as to how it will be able to continue, or are the concerns about smaller sites simply being brushed aside, as my noble friend explicates what the Bill already says?

I will deal with all the points in the speech. If I have not done so by the end, and if my noble friend wants to intervene again, I would be more than happy to hear further questions, either to answer now or write to him about.

Amendments 128 to 133 and 143 to 153, in the names of the right reverend Prelate the Bishop of Derby and the noble Lord, Lord Stevenson of Balmacara, seek to ensure that priority offences relating to modern slavery and human trafficking, where they victimise children, are included in Schedule 6. These amendments also seek to require technology companies to report content which relates to modern slavery and the trafficking of children—including the criminal exploitation of children—irrespective of whether it is sexual exploitation or not. As noble Lords know, the strongest provisions in the Bill relate to children’s safety, and particularly to child sexual exploitation and abuse content. These offences are captured in Schedule 6. The Bill includes a power for Ofcom to issue notices to companies requiring them to use accredited technology or to develop new technology to identify, remove and prevent users encountering such illegal content, whether communicated publicly or privately.

These amendments would give Ofcom the ability to issue such notices for modern slavery content which affects children, even when there is no child sexual exploitation or abuse involved. That would not be appropriate for a number of reasons. The power to tackle illegal content on private communications has been restricted to the identification of content relating to child sexual exploitation and abuse because of the particular risk to children posed by content which is communicated privately. Private spaces online are commonly used by networks of criminals to share illegal images—as we have heard—videos, and tips on the commitment of these abhorrent offences. This is highly unlikely to be reported by other offenders, so it will go undetected if companies do not put in place measures to identify it. Earlier in Committee, the noble Lord, Lord Allan, suggested that those who receive it should report it, but of course, in a criminal context, a criminal recipient would not do that.

Extending this power to cover the identification of modern slavery in content which is communicated privately would be challenging to justify and could represent a disproportionate intrusion into someone’s privacy. Furthermore, modern slavery is usually identified through patterns of behaviour or by individual reporting, rather than through content alone. This reduces the impact that any proactive technology required under this power would have in tackling such content. Schedule 6 already sets out a comprehensive list of offences relating to child sexual exploitation and abuse which companies must tackle. If these offences are linked to modern slavery—for example, if a child victim of these offences has been trafficked—companies must take action. This includes reporting content which amounts to an offence under Schedule 6 to the National Crime Agency or another reporting body outside of the UK.

My noble friend Lord Moylan’s Amendment 135 seeks to remove the offence in Section 5 of the Public Order Act 1986 from the list of priority offences. His amendment would mean that platforms were not required to take proactive measures to reduce the risk of content which is threatening or abusive, and intended to cause a user harassment, alarm or distress, from appearing on their service. Instead, they would be obliged to respond only once they are made aware of the content, which would significantly reduce the impact of the Bill’s framework for tackling such threatening and abusive content. Given the severity of the harm which can be caused by that sort of content, it is right that companies tackle it. Ofcom will have to include the Public Order Act in its guidance about illegal content, as provided for in Clause 171.

Government Amendments 136A to 136C seek to strengthen the illegal content duties by adding further priority offences to Schedule 7. Amendments 136A and 136B will add human trafficking and illegal entry offences to the list of priority offences in the Bill. Crucially, this will mean that platforms will need to take proactive action against content which encourages or assists others to make dangerous, illegal crossings of the English Channel, as well as those who use social media to arrange or facilitate the travel of another person with a view to their exploitation.

The noble Lord, Lord Allan, asked whether these amendments would affect the victims of trafficking themselves. This is not about going after the victims. Amendment 136B addresses only content which seeks to help or encourage the commission of an existing immigration offence; it will have no impact on humanitarian communications. Indeed, to flesh out a bit more detail, Section 2 of the Modern Slavery Act makes it an offence to arrange or facilitate the travel of another person, including through recruitment, with a view to their exploitation. Facilitating a victim’s travel includes recruiting them. This offence largely appears online in the form of advertisements to recruit people into being exploited. Some of the steps that platforms could put in place include setting up trusted flagger programmes, signposting users to support and advice, and blocking known bad actors. Again, I point to some of the work which is already being done by social media companies to help tackle both illegal channel crossings and human trafficking.

Government Amendment 136C will add the offence of foreign interference to the list of priority offences in the Bill. As your Lordships will know, the Government previously made an amendment via the National Security Bill to include this offence in this Bill. Because of the relative pace at which these two Bills are now passing through Parliament, we are now doing it directly in the Online Safety Bill.

My noble friend Lady Buscombe’s Amendment 137 seeks to list the false and threatening communication offences in Schedule 7. Listing the communication offences as priority offences would require platforms to identify and determine the illegality of such content proactively. I appreciate the reasons she set out for raising this issue, but as these offences rely heavily on a user’s mental state, it would be challenging for services to identify this content without significant additional context. Let me reassure her, however, that platforms will still need to have systems and processes in place to remove this content quickly when it is reported to them, as with all other illegal content which is not in Schedule 7.

My noble friend Lord Bethell anticipated later debates on age verification and pornography. If he permits, I will come back on his points then. I have noted his question for that discussion as well as the question from the noble Lord, Lord Stevenson, on financial scams and fraud, which we will have the chance to discuss in full. I am not sure if my noble friend Lord Moylan wants to ask a further question at this juncture or to accept a reassurance that I will consult the Official Report and write on any further points he raised which I have not dealt with.

My Lords, it is genuinely difficult to summarise such a wide-ranging debate, which was of a very high standard. Only one genuinely bright idea has emerged from the whole thing: as we go through Committee, each group of amendments should be introduced by the noble Lord, Lord Allan of Hallam, because it is only after I have heard his contribution on each occasion that I have begun to understand the full complexity of what I have been saying. I suspect I am not alone in that and that we could all benefit from hearing the noble Lord before getting to our feet. That is not meant to sound the slightest bit arch; it is absolutely genuine.

The debate expressed a very wide range of concerns. Concerns about gang grooming and recruiting were expressed on behalf of the right reverend Prelate the Bishop of Derby and my noble friend Lady Buscombe expressed concerns about trolling of country businesses. However, I think it is fair to say that most speakers focused on the following issues. The first was the definition of legality, which was so well explicated by the noble Lord, Lord Allan of Hallam. The second was the judgment bar that providers have to pass to establish whether something should be taken down. The third was the legislative mandating of private foreign companies to censor free speech rights that are so hard-won here in this country. These are the things that mainly concern us.

I was delighted that I found myself agreeing so much with what the noble Baroness, Lady Kidron, said, even though she was speaking in another voice or on behalf of another person. If her own sentiments coincide with the sentiments of the noble Viscount—

I am sorry to intrude, but I must say now on the record that I was speaking on my own behalf. The complication of measuring and those particular things are terribly important to establish, so I am once again happy to agree with the noble Lord.

I am delighted to hear the noble Baroness say that, and it shows that that pool of common ground we share is widening every time we get to our feet. However, the pool is not particularly widening, I am afraid to say—at least in respect of myself; other noble Lords may have been greatly reassured—as regards my noble friend the Minister who, I am afraid, has not in any sense addressed the issues about free speech that I and many other noble Lords raised. On some issues we in the Committee are finding a consensus that is drifting away from the Minister. We probably need to put our heads together more closely on some of these issues with the passage of time in Committee.

My noble friend also did not say anything that satisfied me in respect of the practical operation of these obligations for smaller sites. He speaks smoothly and persuasively of risk-based proactive approaches without saying that, for a large number of sites, this legislation will mean a complete re-engineering of their business model. For example, where Wikipedia operates in a minority language, such as in Welsh Wikipedia, which is the largest Welsh language website in the world, if its model is to involve monitoring what is put out by the community and correcting it as it goes along, rather than having a model in advance that is designed to prevent things being put there in the first place, then it is very likely to close down. If that is one of the consequences of this Bill the Government will soon hear about it.

Finally, although I remain concerned about public order offences, I have to say to the Minister that if he is so concerned about the dissemination of alarm among the population under the provisions of the Public Order Act, what does he think that His Majesty’s Government were doing on Sunday at 3 pm? I beg leave to withdraw the amendment.

Amendment 17 withdrawn.

Amendment 18 not moved.

Amendment 18A

Moved by

18A: Clause 9, page 8, line 23, at end insert—

“(8A) A duty to summarise in the terms of service the findings of the most recent illegal content risk assessment of a service (including as to levels of risk and as to nature, and severity, of potential harm to individuals).”Member’s explanatory statement

This amendment requires providers of Category 1 services to summarise (in their terms of service) the findings of their latest risk assessment regarding illegal content and activity. The limitation to Category 1 services is achieved by an amendment in the name of the Minister to clause 6.

Amendment 18A agreed.

Clause 9, as amended, agreed.

Clause 10: Children’s risk assessment duties

Amendment 19 not moved.

Moved by

20: Clause 10, page 9, line 11, leave out paragraphs (a) to (h) and insert—

“(a) the level of risk that children who are users of the service encounter the harms as outlined in Schedule (Online harms to children) by means of the service;(b) any of the level of risks to children encountered singularly or in combination, having regard to—(i) the design of functionalities, algorithms and other features that present or increase risk of harm, such as low-privacy profile settings by default;(ii) the business model, revenue model, governance, terms of service and other systems and processes or mitigation measures that may reduce or increase the risk of harm;(iii) risks which can build up over time;(iv) the ways in which level of risks can change when experienced in combination with others;(v) the level of risk of harm to children in different age groups;(vi) the level of risk of harm to children with certain characteristics or who are members of certain groups; and(vii) the different ways in which the service is used including but not limited to via virtual and augmented reality technologies, and the impact of such use on the level of risk of harm that might be suffered by children;(c) whether the service has shown regard to the rights of children as set out in the United Nations Convention on the Rights of the Child (see general comment 25 on children’s rights in relation to the digital environment).”Member’s explanatory statement

This amendment would require providers to look at and assess risks on their platform in the round and in line with the 4 Cs of online risks to children (content, contact, conduct and contractual/commercial risks). Although these risks will not be presented on every service, this amendment requires providers to reflect on these risks, so they are not forgotten and can be built into future development of the service.

My Lords, this amendment and Amendments 74, 93 and 123 are part of a larger group that have been submitted as a package loosely referred to as the AV and harms package. They have been the subject of much private debate with the Government, for which we are grateful, and among parliamentarians, and have featured prominently in the media. The amendments are in my name and those of the noble Lord, Lord Bethell, the right reverend Prelate the Bishop of Oxford and the noble Lord, Lord Stevenson, but enjoy the support of a vast array of Members of both Houses. I thank all those who have voiced their support.

The full package of amendments defines and sets out the rules of the road for age assurance, including the timing of its introduction, and the definition of terms such as age verification and age assurance. They introduce the concept of measuring the efficacy of systems with one eye on the future so that we as parliamentarians can indicate where and when we feel that proportionality is appropriate and where it is simply not—for example, in relation to pornography. In parallel, we have developed a schedule of harms, which garners rather fewer column inches but is equally important in establishing Parliament’s intention. It is that schedule of harms that is up for debate today.

Before I lay out the amendment, I thank the 26 children’s charities which have so firmly got behind this package and acknowledge, in particular, Barnardo’s, CEASE and 5Rights, of which I am chair, which have worked tirelessly to ensure that the full expertise of children’s charities has been embedded in these amendments. I also pay tribute to the noble Baroness, Lady Benjamin, who in this area of policy has shown us all the way.

The key amendment in this group is Amendment 93, which would place a schedule of harms to children in the Bill. There are several reasons for doing so, the primary one being that by putting them in the Bill we are stating the intention of Parliament, which gives clarity to companies and underlines the authority of Ofcom to act on these matters. Amendments 20, 74 and 123 ensure that the schedule is mirrored in risk assessments and tasks Ofcom with updating its guidance every six months to capture new and emerging harms, and as such are self-evident.

The proposed harms schedule is centred around the four Cs, a widely used and understood taxonomy of harm used in legislation and regulation around the globe. Importantly, rather than articulate individual harms that may change over time, it sets its sight on categories of harm: content, contact, conduct and contract, which is sometimes referred to as commercial harm. It also accounts for cumulative harms, where two or more risk factors create a harm that is greater than any single harm or is uniquely created by the combination. The Government’s argument against the four Cs is that they are not future-proof, which I find curious since the very structure of the four Cs is to introduce broad categories of harm to which harms can be added, particularly emerging harms. By contrast, the Government are adding an ever-growing list of individual harms.

I wish to make three points in favour of our package of amendments relating first to language, secondly to the nature of the digital world, and finally to clarity of purpose. It is a great weakness of the Bill that it consistently introduces new concepts and language—for example, the terms “primary priority content”, “priority content” and “non-designated content”. These are not terms used in other similar Bills across the globe, they are not evident in current UK law and they do not correlate with established regimes, such as equalities legislation or children’s rights under the convention, more of which in group 7.

The question of language is non-trivial. It is the central concern of those who fight CSAE around the world, who frequently find that enforcement against perpetrators or takedown is blocked by legal systems that define child sexual abuse material differently—not differently in some theoretical sense but because the same image can be categorised differently in two countries and then be a barrier to enforcement across jurisdictions. Leadership from WeProtect, the enforcement community and representatives that I recently met from Africa, South America and Asia have all made this point. It undermines the concept of UK leadership in child protection that we are wilfully and deliberately rejecting accepted language which is embedded in treaties, international agreements and multilateral organisations to start again with our own, very likely with the same confused outcome.

Secondly, I am concerned that while both the Bill and the digital world are predicated on system design, the harms are all articulated as content with insufficient emphasis on systems harms, such as careless recommendations, spreading engagement and the sector-wide focus on maximising engagement, which are the very things that create the toxic and dangerous environment for children. I know, because we have discussed it, that the Minister will say that this is all in the risk assessment, but the risk assessment asks regulated companies to assess how a number of features contribute to harm, mostly expressed as content harm.

What goes through my mind is the spectre of Meta’s legal team, which I watched for several days during Molly Russell’s inquest; they stood in a court of law and insisted that hundreds, in fact thousands, of images of cut bodies and depressive messages did not constitute harm. Rather, they regarded them as cries for help or below the bar of harm as they interpreted it. Similarly, there was material that featured videos of people jumping off buildings—some of them sped-up versions of movie clips edited to suggest that jumping was freedom—and I can imagine a similar argument that says that kind of material cannot be considered harmful, because in another context it is completely legitimate. Yet this material was sent to Molly at scale.

It is not good enough to characterise harms simply by establishing what is or is not harmful content. The previous debate really underlined that it takes a long time and it is very complicated to see what is harmful. But we must make utterly clear that the drip feed of nudges, enticements and recommendations and the creation of a toxic environment, overwhelming a child of 14 with more than 1,400 messages, whether they meet that bar of harmful content or not, is in itself a harm. A jukebox of content harms is not future-proof, and it fails to name the risks of the system. It is to misunderstand where the power of digital design actually lies.

Finally, there is the question of simplicity and clarity. As we discussed on the first day of Committee, business wants clarity, campaigners want clarity, parents want clarity, and Ofcom could do with some clarity. If not the four Cs, my challenge to the Government is to deliver a schedule that has the clarity and simplicity of the amendments in front of us, in which harm is defined by category not by individual content measurements, so that it is flexible now and into the future, and foregrounds the specific role of the system design not only as an accomplice to the named harm but as a harm itself. I beg to move.

My Lords, it is a pleasure to follow the noble Baroness, Lady Kidron. I have listened intently today, and there is no doubt that this Bill not only presents many challenges but throws up the complexity of the whole situation. I think it was the noble Lord, Lord Kamall, in an earlier group who raised the issues of security, safety and freedom. I would add the issue of rights, because we are trying to balance all these issues and characterise them in statute, vis-à-vis the Bill.

On Tuesday, we spoke about one specific harm—pornography—on the group of amendments that I had brought forward. But I made clear at that time that I believe this is not the only harm, and I fully support the principles of the amendments from the noble Baroness, Lady Kidron. I would obviously like to get some clarity from her on the amendments, particularly as to how they relate to other clauses in the Bill.

The noble Baroness has been the pioneer in this field, and her expertise is well recognised across the House. I believe that these amendments really take us to the heart of the Bill and what we are trying to achieve—namely, to identify online harms to children, counteract them and provide a level of safety to young people.

As the noble Lord, Lord Clement-Jones, said on Tuesday,

“there is absolutely no doubt that across the Committee we all have the same intent; how we get there is the issue between us”.—[Official Report, 25/4/23; col. 1196.]

There is actually not that much between us. I fully agree with the principle of putting some of the known harms to children in the Bill. If we know the harms, there is little point in waiting for them to be defined in secondary legislation by Clause 54.

It is clear to me that there are harms to children that we know about, and those harms will not change. It would be best to name those harms clearly in the Bill when it leaves this House. That would allow content providers, search engines and websites in scope of the Bill to prepare to make any changes they need to keep children safe. Perhaps the Minister could comment on that aspect. We also know that parents will expect some harms to be in the Bill. The noble Baroness, Lady Kidron, laid out what they are, and I agree with her analysis. These issues are known and we should not wait for them to be named.

While known harms should be placed into the Bill, I know, understand and appreciate that the Government are concerned about future-proofing. However, I am of the view that a short list of key topics will not undermine that principle. Indeed, the Joint Committee’s report on the draft Bill stated,

“we recommend that key, known risks of harm to children are set out on the face of the Bill”.

In its report on the Bill, the DCMS Select Committee in the other place agreed, saying

“that age-inappropriate or otherwise inherently harmful content and activity (like pornography, violent material, gambling and content that promotes or is instructive in eating disorders, self-harm and suicide) should appear on the face of the Bill”.

Has there been any further progress in discussions on those issues?

At the beginning of the year, the Children’s Commissioner urged Parliamentarians

“to define pornography as a harm to children on the fact of the … Bill, such that the regulator, Ofcom, may implement regulation of platforms hosting adult content as soon as possible following the passage of the Bill”.

I fully agree with the Children’s Commissioner. While the ways in which pornographic content is delivered will change over time, the fact that pornography is harmful to children will not change. Undoubtedly, with the speed of technology—something that the noble Lord, Lord Allan of Hallam, knows a lot more about than the rest of us, having worked in this field—it will no doubt change and we will be presented with new types of challenges.

I therefore urge the Government to support the principle that the key risks are in the Bill, and I thank the noble Baroness, Lady Kidron, for raising this important principle. However, I hope she will indulge me as I seek to probe some of the detail of her amendments and their interactions with the architecture of other parts of the Bill. As I said when speaking to Clause 49 on Tuesday, the devil is obviously in the detail.

First, Clause 54 defines what constitutes

“Content that is harmful to children”,

and Clause 205 defines harm, and Amendment 93 proposes an additional new list of harms. As I have already said, I fully support the principle of harms being in the Bill, but I raise a question for the noble Baroness. How does she see these three definitions working together? That might refer back to a preliminary discussion that we had in the tearoom earlier.

These definitions of harms are in addition to the content to be defined as primary priority content and priority content. Duties in Clauses 11 and 25 continue to refer to these two types of content for Part 3 services, but Amendments 20 and 74 would remove the need for risk assessments in Clauses 10 and 24 to address these two types of content. It seems that the amendments could create a tension in the Bill, and I am interested to ascertain how the noble Baroness, Lady Kidron, foresees that tension operating. Maybe she could give us some detail in her wind-up about that issue. An explanation of that point may bring some clarity to understanding how the new schedule that the noble Baroness proposes will work alongside the primary priority content and the priority content lists. Will the schedule complement primary priority content, or will it be an alternative?

Secondly, as I said, some harms are known but there are harms that are as yet unknown. Will the noble Baroness, Lady Kidron, consider a function to add to the list of content in her Amendment 93, in advance of us coming back on Report? There is no doubt that the online space is rapidly changing, as this debate has highlighted. I can foresee a time when other examples of harm should be added to the Bill. I accept that the drafting is clear that the list is not exclusive, but it is intended to be a significant guide to what matters to the public and Parliament. I also accept that Ofcom can provide guidance on other content under Amendment 123, but, without a regulatory power added to Amendment 93, it feels that we are perhaps missing a belt-and-braces approach to online harms to children. After all, our principal purpose here is to protect children from online harm.

I commend the noble Baroness, Lady Kidron, on putting these important amendments before the Committee, and I fully support the principle of what she seeks to achieve. But I hope that, on further reflection, she will look at the points I have suggested. Perhaps she might suggest other ideas in her wind-up, and we could have further discussions in advance of Report. I also look forward to the Minister’s comments on these issues.

My Lords, I support Amendments 20, 93 and 123, in my name and those of the noble Baroness, Lady Kidron, and the noble Lords, Lord Bethell and Lord Stevenson. I also support Amendment 74 in the name of the noble Baroness, Lady Kidron. I pay tribute to the courage of all noble Lords and their teams, and of the Minister and the Bill team, for their work on this part of the Bill. This work involves the courage to dare to look at some very difficult material that, sadly, shapes the everyday life of too many young people. This group of amendments is part of a package of measures to strengthen the protections for children in the Bill by introducing a new schedule of harms to children and plugging a chronological gap between Part 3 and Part 5 services, on when protection from pornography comes into effect.

Every so often in these debates, we have been reminded of the connection with real lives and people. Yesterday evening, I spent some time speaking on the telephone with Amanda and Stuart Stephens, the mum and dad of Olly Stephens, who lived in Reading, which is part of the diocese of Oxford. Noble Lords will remember that Olly was tragically murdered, aged 13, in a park near his home, by teenagers of a similar age. Social media played a significant part in the investigation and in the lives of Olly and his friends—specifically, social media posts normalising knife crime and violence, with such a deeply tragic outcome.

Last year in June, “Panorama” dared to look into this world. The programme revealed the depth and extent of the normalisation of knives and knife crime in posts offered to young people. I was struck by the comments of Frances Haugen, filmed when she met Stuart and Amanda. She said that each of us sees social media through a pinhole: a tiny snapshot of the total content. We have no idea how much darkness and evil are shaping children and young people, destroying their sense of proportion and deeply affecting offline behaviour. The only group that has the whole picture, of course, is the companies themselves.

The noble Baroness, Lady Kidron, and others have outlined the remarkable degree of support for this raft of amendments from charities working to protect children. We should listen. These amendments will ensure a much wider definition of “harm” and will again future-proof the Bill in terms of technology which is even now coming over the horizon.

The Center for Countering Digital Hate speaks about an arms race to devise ever more effective ways of keeping users’ attention, even if it means putting them at risk. Its researchers set up new accounts in the United States, United Kingdom, Canada and Australia at the minimum age TikTok allows: 13 years old. Those accounts paused briefly on videos about body image and mental health and liked them. What the researchers found was deeply disturbing. Within 2.6 minutes, TikTok recommended suicide content. Within 8 minutes, TikTok served content relating to eating disorders. Every 39 seconds, TikTok recommended videos about body image and mental health to teens. CCDH researchers found a community for eating disorder content on the platform amassing 13.2 billion views across 56 hashtags, often designed to evade moderation.

As the noble Baroness, Lady Kidron, said, this fourfold classification of harms to children is being adapted elsewhere in the world, including the European Union. The schedule in the amendment gives clear but non-exhaustive examples to guide service providers on the meaning of each of the four Cs. It is vital to have more comprehensive agreed definitions of harm in the Bill.

I will reflect for a moment on what each of the four Cs means. Content harms are the most familiar. At the moment, children who go online are likely to encounter age-inappropriate content, including violent, gory and graphic communication, hate speech, terrorism, online prostitution, drugs, eating disorders and self-harm. Research also shows that exposure to different types of harmful content is interrelated: so, if a child reports seeing one type of disturbing content, it is likely that they have seen others as well.

Secondly, contact harms encourage harmful actions in the non-virtual world. A 10 year-old girl was left with burns after spraying an aerosol deodorant with the nozzle right up against her skin to create a freezing sensation. Jane Platt’s daughter Sarah, aged 15, was rushed to hospital in February 2020 after doing the “skull-breaker challenge”, which involves two people kicking the legs from under a third, making them fall over. These suggestions could never be offered in young people’s magazines or broadcast media.

Thirdly, there are conduct harms. In a global survey, 54% of young people—57% of girls and 48% of boys—reported having experienced online sexual harms before they were 18 years old, including within interaction with adults and being asked something sexually explicit or being sent sexually explicit content.

Finally, there are commercial harms. Over half of the games on Google Play now include loot boxes and more than 93% of games that feature loot boxes are marked suitable for children aged 12 years-plus.

As the noble Baroness, Lady Kidron, and others have argued, these harms are often cumulative and interrelated. The social media companies are the only ones not looking through a keyhole but monitoring social media in the round and able to assess what is happening, but evidence suggests that they will do not so until compelled by legislation. These amendments are a vital step forward in fulfilling the Bill’s purpose of providing additional protection from harm for children. I urge the Government to adopt them.

My Lords, I really appreciated the contribution from the noble Baroness, Lady Ritchie of Downpatrick, because she asked a lot of questions about this group of amendments. Although I might be motivated by different reasons, I found it difficult to fully understand the impact of the amendments, so I too want to ask a set of questions.

Harm is defined in the Bill as “physical or psychological harm”, and there is no further explanation. I can understand the frustration with that and the attempts therefore to use what are described as the

“widely understood and used 4 Cs of online risk to children”.

They are not widely understood by me, and I have ploughed my way through it. I might well have misunderstood lots in it, but I want to look at and perhaps challenge some of the contents.

I was glad that Amendment 20 recognises the level of risk of harm to different age groups. That concerns me all the time when we talk about children and young people, and then end up treating four year-olds, 14 year-olds and 18 year-olds. I am glad that that is there, and I hope that we will look at it again in future.

I want to concentrate on Amendment 93 and reflect and comment more generally on the problem of a definition, or a lack of definition, of harm in the Bill. For the last several years that we have been considering bringing this Bill to this House and to Parliament, I have been worried about the definition of psychological harm. That is largely because this category has become ever more expansive and quite subjective in our therapeutic age. It is a matter of some discussion and quite detailed work by psychologists and professionals, who worry that there is an expanding concept group of what is considered harmful and what psychological harm really means.

As an illustration, I was invited recently to speak to a group of sixth-formers and was discussing things such as trigger warnings and so on. They said, “Well, you know, you’ve got to understand what it’s like”—they were 16 year-olds. “When we encounter certain material, it makes us have PTSD”. I was thinking, “No, it doesn’t really, does it?” Post-traumatic stress disorder is something that you might well gain if you have been in the middle of a war zone. The whole concept of triggering came from psychological and medical insights from the First World War, which you can understand. If you hear a car backfiring, you think it is somebody shooting at you. But the idea here is that we should have trigger warnings on great works of literature and that if we do not it will lead to PTSD.

I am not being glib, because an expanded, elastic and pathologised view of harm is being used quite cavalierly and casually in relation to young people and protecting them, often by the young people themselves. It is routinely used to close down speech as part of the cancel culture wars, which, as noble Lords know, I am interested in. Is there not a danger that this concept of harm is not as obvious as we think, and that the psychological harm issue makes it even more complicated?

The other thing is that Amendment 93 says:

“The harms in this Schedule are a non-exhaustive list of categories and other categories may be relevant”.

As with the discussion on whose judgment decides the threshold for removing illegal material, I think that judging what is harmful is even more tricky for the young in relation to psychological harm. I was reminded of that when the noble Baroness, Lady Kidron, complained that what she considered to be obviously and self-evidently harmful, Meta did not. I wondered whether that is just the case with Meta, or whether views will differ when it comes to—

The report found—I will not give a direct quotation—that social media contributed to the death of Molly Russell, so it was the court’s judgment, not mine, that Meta’s position was indefensible.

I completely understand that; I was making the point that there will be disagreements in judgments. In that instance, it was resolved by a court, but we are talking about a situation where I am not sure how the judgment is made.

In these amendments, there are lists of particular harms—a variety are named, including self-harm—and I wanted to provide some counterexamples of what I consider to be harms. I have been inundated by algorithmic adverts for “Naked Education” on Channel 4, maybe because of the algorithms I am on. I think that the programme is irresponsible; I say that having watched it, rather than just having read a headline. Channel 4 is posing this programme with naked adults and children as educational by saying that it is introducing children to the naked body. I think it is harmful for children and that it should not be on the television, but it is advertised on social media—I have seen quite a lot of it.

The greatest example of self-harm we encounter at present is when gender dysphoric teenagers—as well as some younger than teenagers; they are predominately young women—are affirmed by adults, as a kind of social contagion, into taking body-changing and body-damaging hormones and performing self-mutilation, whether by breast binding or double mastectomies, which is advertised and praised by adults. That is incredibly harmful for young people, and it is reflected online at lot, because much of this is discussed, advertised or promoted online.

This is related to the earlier contributions, because I am asking: should those be added to the list of obvious harms? Although not many noble Lords are in the House now, if there were many more here, they would object to what I am saying by stating, “That is not harmful at all. What is harmful is what you’re saying, Baroness Fox, because you’re causing psychological harm to all those young people by being transphobic”. I am raising these matters because we think we all agree that there is a consensus on what is harmful material online for young people, but it is not that straightforward.

The amendment states that the Bill should target any platform that posts

“links to, or … encourages child users to seek”

out “dangerous or illegal activity”. I understand “illegal activity”, but on “dangerous” activities, I assume that we do not mean extreme sports, mountain climbing and so on, which are dangerous—that comes to mind probably because I have spent too much time with young people who spend their whole time looking at those things. I worry about the unintended consequences of things being banned or misinterpreted in that way.

To respond briefly to the noble Baroness, I shall give a specific example of how Amendment 93 would help. Let us go back to the coroner’s courtroom where the parents of Molly Russell were trying to get the coroner to understand what had happened to their daughter. The legal team from Meta was there, with combined salaries probably in seven figures, and the argument was about the detail of the content. At one point, I recall Ian Russell saying that one of the Meta lawyers said, “We are topic agnostic”. I put it to the noble Baroness that, had the provisions in Amendment 93 been in place, first, under “Content harms” in proposed new paragraph 3(c) and (d), Meta would have been at fault; under “Contact harms” in proposed new paragraph 4(b), Meta would have been at fault; under “Conduct harms” in proposed new paragraph 5(b), Meta would have been at fault; and under “Commercial harms” in proposed new paragraph 6(a) and (b), Meta would have been at fault. That would have made things a great deal simpler.

I appreciate that that this is the case we all have in the back of our minds. I am asking whether, when Meta says it is content agnostic, the Bill is the appropriate place for us to list the topics that we consider harmful. If we are to do that, I was giving examples of contentious, harmful topics. I might have got this wrong—

I will answer the noble Baroness more completely when I wind up, but I just want to say that she is missing the point of the schedule a little. Like her, I am concerned about the way we concentrate on content harms, but she is bringing it back to content harms. If she looks at it carefully, a lot of the provisions are about contact and conduct: it is about how the system is pushing children to do certain things and pushing them to certain places. It is about how things come together, and I think she is missing the point by keeping going back to individual pieces of content. I do not want to take the place of the Minister, but this is a systems and processes Bill; it is not going to deal with individual pieces of content in that way. It asks, “Are you creating these toxic environments for children? Are you delivering this at scale?” and that is the way we must look at this amendment.

I will finish here, because we have to get on, but I did not introduce content; it is in the four Cs. One of the four Cs is “content” and I am reacting to amendments tabled by the noble Baroness. I do not think I am harping on about content; I was responding to amendments in which content was one of the key elements.

My Lords, I speak in support of these amendments with hope in my heart. I thank the noble Baroness, Lady Kidron, and the noble Lord, Lord Bethell, for leading the charge with such vigour, passion and determination: I am with them all the way.

The Government have said that the purpose of the Bill is to protect children, and it rests on our shoulders to make sure it delivers on this mission. Last week, on the first day in Committee, the Minister said:

“Through their duties of care, all platforms will be required proactively to identify and manage risk factors associated with their services in order to ensure that users do not encounter illegal content and that children are protected from harmful content. To achieve this, they will need to design their services to reduce the risk of harmful content or activity occurring and take swift action if it does.—[Official Report, 19/4/23; cols. 274-75.]

This is excellent and I thank the Government for saying it. But the full range of harms and risk to children will not be mitigated by services if they do not know what they are expected to risk-assess for and if they must wait for secondary legislation for this guidance.

The comprehensive range of harms children face every day is not reflected in the Bill. This includes sexual content that does not meet the threshold of pornography. This was highlighted recently in an investigation into TikTok by the Telegraph, which found that a 13 year-old boy was recommended a video about the top 10 porn-making countries, and that a 13 year-old girl was shown a livestream of a pornography actor in her underwear answering questions from viewers. This content is being marketed to children without a user even seeking out pornographic content, but this would still be allowed under the Bill.

Furthermore, high-risk challenges, such as the Benadryl and blackout challenges, which encourage dangerous behaviour on TikTok, are not dealt with in the Bill. Some features, such as the ability of children to share their location, are not dealt with either. I declare an interest as vice-president of Barnardo’s, which has highlighted how these features can be exploited by organised criminal gangs that sexually exploit children to keep tabs on them and trap them in a cycle of exploitation.

It cannot be right that the user-empowerment duties in the Bill include a list of harmful content that services must enable adults to toggle off, yet the Government refuse to produce this list for children. Instead, we have to wait for secondary legislation to outline harms to children, causing further delay to the enforcement of services’ safety duties. Perhaps the Minister can explain why this is.

The four Cs framework of harm, as set out in these amendments, is a robust framework that will ensure service risk assessments consider the full range of harms children face. I will repeat it once again: childhood lasts a lifetime, so we cannot fail children any longer. Protections are needed now, not in years to come. We have waited far too long for this. Protections need to be fast-tracked and must be included in the Bill. That is why I fully support these amendments.

My Lords, in keeping with the Stevenson-Knight double act, I am leaving it to my noble friend to wind up the debate. I will come in at this point with a couple of questions and allow the Minister to have a bit of time to reflect on them. In doing so, I reinforce my support for Amendment 295 in the name of the noble Lord, Lord Russell, which refers to volume and frequency also being risk factors.

When I compare Amendment 20 with Clause 10(6), which refers to children’s risk assessments and what factors should be taken into account in terms of the risk profile, I see some commonality and then some further things which Amendment 20, tabled by the noble Baroness, Lady Kidron, adds. In my opinion, it adds value. I am interested in how the Minister sees the Bill, as it stands currently, covering some issues that I will briefly set out. I think it would be helpful if the Committee could understand that there may be ways that the Bill already deals with some of the issues so wonderfully raised by the noble Baroness; it would be helpful if we can flush those out.

I do not see proposed new subsection (b)(iii),

“risks which can build up over time”,

mentioned in the Bill, nor explicit mention of proposed new subsection (b)(iv),

“the ways in which level of risks can change when experienced in combination with others”,

which I think is critical in terms of the way the systems work. Furthermore, proposed new subsection (b)(vii),

“the different ways in which the service is used including but not limited to via virtual and augmented reality technologies”,

starts to anticipate some other potential harms that may be coming very rapidly towards us and our children. Again, I do not quite see it included. I see “the design of functionalities”, “the business model” and “the revenue model”. There is a lot about content in the original wording of the Bill, which is less so here, and, clearly, I do not see anything in respect of the UN Convention on the Rights of the Child, which has been debated in separate amendments anyway. I wanted to give the Minister some opportunity on that.

My Lords, I restate my commitment to Amendments 20, 93 and 123, which are in my name and those of the noble Baroness, Lady Kidron, the right reverend Prelate the Bishop of Oxford, and the noble Lord, Lord Stevenson, and the noble Baroness’s Amendment 74. It is a great honour to follow the noble Lord, Lord Knight. He put extremely well some key points about where there are gaps in the existing Bill. I will build on why we have brought forward these amendments in order to plug these gaps.

In doing so, I wish to say that it has been a privilege to work with the right reverend Prelate, the noble Baroness and the noble Lord, Lord Stevenson. We are not from the same political geographies, but that collaboration demonstrates the breadth of the political concern, and the strength of feeling across the Committee, about these important gaps when it comes to harms—gaps that, if not addressed, will put children at great risk. In this matter we are very strongly united. We have been through a lot together, and I believe this unlikely coalition demonstrates how powerful the feelings are.

It has been said before that children are spending an increasing amount of their lives online. However, the degree of that inflection point in the last few years has been understated, as has how much further it has got to go. The penetration of mobile phones is already around 75% of 10 year-olds—it is getting younger, and it is getting broader.

In fact, the digital world is totally inescapable in the life of a child, whether that is for a young child who is four to six years old or an older child who is 16 or 17. It is increasingly where they receive their education—I do not think that is necessarily a good thing, but that is arguable—it is where they establish and maintain their personal relationships and it is a key forum for their self-expression.

For anyone who suspects otherwise, I wish to make it clear that I firmly believe in innovation and progress, and I regard the benefits of the digital world as really positive. I would never wish to prevent children accessing the benefits of the internet, the space it creates for learning and building community, and the opportunities it opens for them. However, environments matter. The digital world is not some noble wilderness free from original sin or a perfect, frictionless marketplace where the best, nicest, and most beautiful ideas triumph. It is a highly curated experience defined by the algorithms and service agreements of the internet companies. That is why we need rules to ensure that it is a safe space for children.

I started working on my first internet business in 1995, nearly 30 years ago. I was running the Ministry of Sound, and we immediately realised that the internet was an amazing way of getting through to young people. Our target audiences were either clubbers aged over 18 or the younger brothers and sisters of clubbers who bought our merchandise. The internet gave us an opportunity to get past all the normal barriers—past parents and regulation to reach a wonderful new market. I built a good business and it worked out well for me, but those were the days before GDPR and what we understand from the internet. I know from my experience that we need to ensure that children are protected and shielded from the harms that bombard them, because there are strong incentives—mainly financial but also other, malign incentives—for bad actors to use the internet to get through to children.

Unfortunately, as the noble Baroness, Lady Kidron, pointed out, the Bill as it stands does not achieve that aim. Take, for example, contact harms, such as grooming and child sexual abuse. In February 2020, Bark, a US-based organisation that helps families manage and protect their children’s digital lives, launched an 11 year-old persona online who it called Bailey. Bailey’s online persona clearly shows that she is an ordinary 11 year-old, posting content that is ordinary for an 11 year-old. Within 30 seconds of her persona being launched online she received a like from a man whose profile picture was a penis. Within two minutes, multiple messages were received from men, and within five minutes a video call. Shortly afterwards, she received requests from men to meet up. I remind your Lordships that Bailey was 11 years old. These are not trivial content harms; these are attempts to contact a minor using the internet as a medium.

The Bill does try to address contact harms. I am supportive of the Bill and its principles, and am a big fan of the team that is trying to drive it through Parliament. For example, the Bill specifies that features that enable adults to search for and contact children must be risk-assessed. However, that is where the Bill currently stops. There is no comprehensive list of harms or features to inform that risk assessment. For instance, a feature such as live-streaming, which can enable adults to access children directly, is not specifically referenced, meaning that there is no explicit obligation for services to risk-assess such features—a big gap. Services cannot be expected to read between the lines of what the Government’s intentions may be. We must be explicit and clear in the Bill if we are serious about delivering on children’s safety.

That is why our amendments would do four things. First, they would introduce into the Bill a new schedule of harms to children, framed around the four categories of risk: content, contact, conduct and contract or commercial. This will ensure that harms to children in the Bill reflect the full range of harms that children encounter online, including from design features that facilitate pathways to harm to pornography, self-harm, pro-suicide content and grooming. Secondly, we seek to ensure that services must risk-assess for the harms listed in this proposed new schedule in order to ensure that these risk assessments are comprehensive. Thirdly, we seek to task Ofcom with producing guidance, to be updated every 12 months, on new and emerging harms. Fourthly, we seek to ensure that Ofcom consults with children’s advocates and charities when producing this guidance.

The timing of this is very important. The real-life Bailey cannot wait for harms to be outlined in secondary legislation. These problems have been around for a long time but, as I said, the inflection curve on technology is shooting right up the hockey stick at the moment. If the primary purpose of the Bill is to protect children, it must include the harms that children face every day in primary legislation—not in secondary legislation, not after Royal Assent, not in a year or two, but now. Our Amendment 93 would introduce in the Bill a schedule of harms to children. This non-exhaustive list would produce a robust and future-proof framework, and have the clarity and mandate to produce comprehensive risk assessments to ensure that the Bill delivers on its chief purpose: protecting children.

My Lords, I support this group of amendments, so ably introduced by my noble friend and other noble Lords this afternoon.

I am not a lawyer and I would not say that I am particularly experienced in this business of legislating. I found this issue incredibly confusing. I hugely appreciate the briefings and discussions—I feel very privileged to have been included in them—with my noble friend the Minister, officials and the Secretary of State herself in their attempt to explain to a group of us why these amendments are not necessary. I was so determined to try to understand this properly that, yesterday, when I was due to travel to Surrey, I took all my papers with me. I got on the train at Waterloo and started to work my way through the main challenges that officials had presented.

The first challenge was that, fundamentally, these amendments cut across the Bill’s definitions of “primary priority content” and “priority content”. I tried to find them in the Bill. Unfortunately, in Clause 54, there is a definition of primary priority content. It says that, basically, primary priority content is what the Secretary of State says it is, and that content that is harmful to children is primary priority content. So I was none the wiser on Clause 54.

One of the further challenges that officials have given us is that apparently we, as a group of noble Lords, were confusing the difference between harm and risk. I then turned to Clause 205, which comes out with the priceless statement that a risk of harm should be read as a reference to harm—so maybe they are the same thing. I am still none the wiser.

Yesterday morning, I found myself playing what I can only describe as a parliamentary game of Mornington Crescent, as I went round and round in circles. Unfortunately, it was such a confusing game of Mornington Crescent that I forgot that I needed to change trains, ended up in Richmond instead of Redhill, and missed my meeting entirely. I am telling the Committee this story because, as the debate has shown, it is so important that we put in the Bill a definition of the harms that we are intending to legislate for.

I want to address the points made by the noble Baroness, Lady Fox. She said that we might not all agree on what harms are genuinely harmful for children. That is precisely why Parliament needs to decide this, rather than abdicate it to a regulator who, as other noble Lords said earlier today, is then put into a political space. It is the job of Parliament to decide what is dangerous for our children and what is not. That is the approach that we take in the physical world, and it should be the approach that we take in the online world. We should do that in broad categories, which is why the four Cs is such a powerful framework. I know that we are all attempting to predict the known unknowns, which is impossible, but this framework, which gives categories of harm, is clear that it can be updated, developed and, as my noble friend Lord Bethell, said, properly consulted on. We as parliamentarians should decide; that is the purpose of voting in Parliament.

I have a couple of questions for my noble friend the Minister. Does he agree that Parliament needs to decide what the categories of online harms are that the Bill is attempting to protect our children from? If he does, why is it not the four Cs? If he really thinks it is not the four Cs, will he bring back an alternative schedule of harms?

My Lords, I will echo the sentiments of the noble Baroness, Lady Harding, in my contribution to another very useful debate, which has brought to mind the good debate that we had on the first day in Committee, in response to the amendment tabled by the noble Lord, Lord Stevenson, in which we were seeking to get into the Bill what we are actually trying to do.

I thought that the noble Baroness, Lady Fox, was also welcoming additional clarity, specifically in the area of psychological harm, which I agree with. Certainly in its earlier incarnations, the Bill was scattered throughout with references, some of which have been removed, but they are very much open to interpretation. I hope that we will come back to that.

I was struck by the point made by the noble Lord, Lord Russell, around what took place in that coroner’s hearing. You had two different platforms with different interpretations of what they thought that their duty of care would be. That is very much the point. In my experience, platforms will follow what they are told to follow. The challenge is when each of them comes to their own individual view around what are often complex areas. There we saw platforms presenting different views about their risk assessments. If we clarify that for them through amendments such as these, we are doing everyone a favour.

I again compliment my noble friend Lady Benjamin for her work in this area. Her speech was also a model of clarity. If we can bring some of that clarity to the legislation and to explaining what we want, that will be an enormous service.

The noble Lord, Lord Knight, made some interesting points around how this would add value to the Bill, teasing out some of the specific gaps that we have there. I look forward to hearing the response on that.

I was interested in the comments from the noble Lord, Lord Bethell, on mobile phone penetration. We should all hold in common that we are not going back to a time BC—before connection. Our children will be connected, which creates the imperative for us to get this right. There has perhaps been a tendency for us to bury our heads in the sand, and occasionally you hear that still—it is almost as if we would wish this world away. However, the noble Baroness, Lady Kidron, is at the other end of the spectrum; she has come alive on this subject, precisely because she recognises that that will not happen. We are in a world where our children will be connected, so it is on us to figure out how we want those connections to work and to instruct the people who provide those connective services on what they should do. It is certainly not for us to imagine that somehow they will all go away. We will come to that in later groups when we talk about minimum ages; if younger children are online, there is a real issue around how we are going to deal with that.

The right reverend Prelate the Bishop of Oxford highlighted some really important challenges based on real experiences that families today are suffering—let us use the word as it should be—and made the case for clarity. I do not know how much we are allowed to talk in praise of EU legislation, but I am looking at the Digital Services Act—I have looked at a lot of EU legislation—and this Bill, and there is a certain clarity to EU regulation, particularly the process of adding recitals, which are attached to the law and explain what it is meant to do. That is sometimes missing here. I know that there are different legal traditions, but you can sometimes look at an EU regulation and the UK law and the former appears to be much clearer in its intent.

That brings me to the substance of my comments in response to this group, so ably introduced by the noble Baroness, Lady Kidron. I hope that the Government heed and recognise that, at present, no ordinary person can know what is happening in the Bill—other than, perhaps, the wife of the noble Lord, Lord Stevenson, who will read it for fun—and what we intend to do.

I was thinking back to the “2B or not 2B” debate we had earlier about the lack of clarity around something even as simple as the classification of services. I was also thinking that, if you ask what the Online Safety Bill does to restrict self-harm content, the answer would be this: if it is a small social media platform, it will probably be categorised as a 2B service, then we can look at Schedule 7, where it is prohibited from assisting suicide, but we might want to come back to some of the earlier clauses with the specific duties—and it will go on and on. As the noble Baroness, Lady Harding, described, you are leaping backwards and forwards in the Bill to try to understand what we are trying to do with the legislation. I think that is a genuine problem.

In effect, the Bill is Parliament setting out the terms of service for how we want Ofcom to regulate online services. We debated terms of service earlier. What is sauce for the goose is sauce for the gander. We are currently failing our own tests of simplicity and clarity on the terms of service that we will give to Ofcom.

As well as platforms, if ordinary people want to find out what is happening, then, just like those platforms with the terms of service, we are going to make them read hundreds of pages before they find out what this legislation is intended to do. We can and should make this simpler for children and parents. I was able to meet Ian Russell briefly at the end of our Second Reading debate. He has been an incredibly powerful and pragmatic voice on this. He is asking for reasonable things. I would love to be able to give a Bill to Ian Russell, and the other families that the right reverend Prelate the Bishop of Oxford referred to, that they can read and that tells them very clearly how Parliament has responded to their concerns. I think we are a long way short of that simple clarity today.

It would be extraordinarily important for service providers, as I already mentioned in response to the noble Lord, Lord Russell. They need that clarity, and we want to make sure that they have no reason to say, “I did not understand what I was being asked to do”. That should be from the biggest to the smallest, as the noble Lord, Lord Moylan, keeps rightly raising with us. Any small service provider should be able to very clearly and simply understand what we are intending to do, and putting more text into the Bill that does that would actually improve it. This is not about adding a whole load of new complications and the bells and whistles we have described but about providing clarity on our intention. Small service providers would benefit from that clarity.

The noble Baroness, Lady Ritchie, rightly raised the issue of the speed of the development of technology. Again, we do not want the small service provider in particular to think it has to go back and do a whole new legal review every time the technology changes. If we have a clear set of principles, it is much quicker and simpler for it to say, “I have developed a new feature. How does it match up against this list?”, rather than having to go to Clause 12, Clause 86, Clause 94 and backwards and forwards within the Bill.

It will be extraordinarily helpful for enforcement bodies such as Ofcom to have a yardstick—again, this takes us back to our debate on the first day—for its prioritisation, because it will have to prioritise. It will not be able to do everything, everywhere, all at once. If we put that prioritisation into the legislation, it will, frankly, save potential arguments between Parliament, the Government and Ofcom later on, when they have decided to prioritise X and we wanted them to prioritise Y. Let us all get aligned on what we are asking them to do up front.

Dare I say—the noble Baroness, Lady Harding, reminded me of this—that it may also be extraordinarily helpful for us as politicians so that we can understand the state of the law. I mean not just the people who are existing specialists or are becoming specialists in this area and taking part in this debate but the other hundreds of Members of both Houses, because this is interesting to everyone. I have experience of being in the other place, and every Member of the other place will have constituents coming to them, often with very tragic circumstances, and asking what Parliament has done. Again, if they have the Online Safety Bill as currently drafted, I think it is hard for any Member of Parliament to be able to say clearly, “This is what we have done”. With those words and that encouraging wind, I hope the Government are able to explain, if not in this way, that they have a commitment to ensuring that we have that clarity for everybody involved in this process.

My Lords, over the last few hours I have praised us for having developed a style of discussion and debate that is certainly relatively new and not often seen in the House, where we have tried to reach out to each other and find common ground. That was not a problem in this last group of just over an hour; I think we are united around the themes that were so brilliantly introduced in a very concise and well-balanced speech by the noble Baroness, Lady Kidron, who has been a leading and inspirational force behind this activity for so long.

Although different voices have come in at different times and asked questions that still need to be answered, I sense that we have reached a point in our thinking, if not in our actual debates, where we need a plan. I too reached this point; that was exactly the motivation I had in tabling Amendment 1, which was discussed on the first day. Fine as the Bill is—it is a very impressive piece of work in every way—it lacks what we need as a Parliament to convince others that we have understood the issues and have the answers to their questions about what this Government, or this country as a whole, are going to do about this tsunami of difference, which has arrived in the wake of the social media companies and search engines, in the way we do our business and live our lives these days. There is consensus, but it is slightly different to the consensus we had in earlier debates, where we were reassuring ourselves about the issues we were talking about but were not reaching out to the Government to change anything so much as being happy that we were speaking the same language and that they were in the same place as we are gradually coming to as a group, in a way.

Just before we came back in after the lunch break, I happened to talk to the noble Lord, Lord Grade, who is the chair of Ofcom and is listening to most of our debates and discussions when his other duties allow. I asked him what he thought about it, and he said that it was fascinating for him to recognise the level of expertise and knowledge that was growing up in the House, and that it would be a useful resource for Ofcom in the future. He was very impressed by the way in which everyone was engaging and not getting stuck in the niceties of the legislation, which he admitted he was experiencing himself. I say that softly; I do not want to embarrass him in any way because he is an honourable man. However, the point he makes is really important.

I say to the Minister that I do not think we are very far apart on this. He knows that, because we have discussed it at some length over the last six to eight weeks. What I think he should take away from this debate is that this is a point where a decision has to be taken about whether the Government are going to go with the consensus view being expressed here and put deliberately into the Bill a repetitive statement, but one that is clear and unambiguous, about the intention behind the Government’s reason for bringing forward the Bill and for us, the Opposition and other Members of this House, supporting it, which is that we want a safe internet for our children. The way we are going to do that is by having in place, up front and clearly in one place, the things that matter when the regulatory structure sits in place and has to deal with the world as it is, of companies with business plans and business models that are at variance with what we think should be happening and that we know are destroying the lives of people we love and the future of our country—our children—in a way that is quite unacceptable when you analyse it down to its last detail.

It is not a question of saying back to us across the Dispatch Box—I know he wants to but I hope he will not—“Everything that you have said is in the Bill; we don’t need to go down this route, we don’t need another piece of writing that says it all”. I want him to forget that and say that actually it will be worth it, because we will have written something very special for the world to look at and admire. It is probably not in its perfect form yet, but that is what the Government can do: take a rough and ready potential diamond, polish it, chamfer it, and bring it back and set it in a diadem we would all be proud to wear—Coronations excepted—so that we can say, “Look, we have done the dirty work here. We’ve been right down to the bottom and thought about it. We’ve looked at stuff that we never thought in our lives we would ever want to see and survived”.

I shake at some of the material we were shown that Molly Russell was looking at. But I never want to be in a situation where I will have to say to my children and grandchildren, “We had the chance to get this right and we relied on a wonderful piece of work called the Online Safety Act 2023; you will find it in there, but it is going to take you several weeks and a lot of mental harm and difficulty to understand what it means”.

So, let us make it right. Let us not just say “It’ll be alright on the night”. Let us have it there. It is almost right but, as my noble friend Lord Knight said, it needs to be patched back into what is already in the Bill. Somebody needs to look at it and say, “What, out of that, will work as a statement to the world that we care about our kids in a way that will really make a difference?” I warn the Minister that, although I said at Second Reading that I wanted to see this Bill on the statute book as quickly as possible, I will not accept a situation where we do not have more on this issue.

I am grateful to all noble Lords who have spoken on this group and for the clarity with which the noble Lord, Lord Stevenson, has concluded his remarks.

Amendments 20, 74, 93 and 123, tabled by the noble Baroness, Lady Kidron, would mean a significant revising of the Bill’s approach to content that is harmful to children. It would set a new schedule of harmful content and risk to children—the 4 Cs—on the face of the Bill and revise the criteria for user-to-user and search services carrying out child safety risk assessments.

I start by thanking the noble Baroness publicly—I have done so privately in our discussions—for her extensive engagement with the Government on these issues over recent weeks, along with my noble friends Lord Bethell and Lady Harding of Winscombe. I apologise that it has involved the noble Baroness, Lady Harding, missing her stop on the train. A previous discussion we had also very nearly delayed her mounting a horse, so I can tell your Lordships how she has devoted hours to this—as they all have over recent weeks. I would like to acknowledge their campaigning and the work of all organisations that the noble Baroness, Lady Kidron, listed at the start of her speech, as well as the families of people such as Olly Stephens and the many others that the right reverend Prelate the Bishop of Oxford mentioned.

I also reassure your Lordships that, in developing this legislation, the Government carried out extensive research and engagement with a wide range of interested parties. That included reviewing international best practice. We want this to be world-leading legislation, including the four Cs framework on the online risks of harm to children. The Government share the objectives that all noble Lords have echoed in making sure that children are protected from harm online. I was grateful to the noble Baroness, Lady Benjamin, for echoing the remarks I made earlier in Committee on this. I am glad we are on the same page, even if we are still looking at points of detail, as we should be.

As the noble Baroness, Lady Kidron, knows, it is the Government’s considered opinion that the Bill’s provisions already deliver these objectives. I know that she remains to be convinced, but I am grateful to her for our continuing discussions on that point, and for continuing to kick the tyres on this to make sure that this is indeed legislation of which we can be proud.

It is also clear that there is broad agreement across the House that the Bill should tackle harmful content to children such as content that promotes eating disorders, illegal behaviour such as grooming and risk factors for harm such as the method by which content is disseminated, and the frequency of alerts. I am pleased to be able to put on record that the Bill as drafted already does this in the Government’s opinion, and reflects the principles of the four Cs framework, covering each of those: content, conduct, contact and commercial or contract risks to children.

First, it is important to understand how the Bill defines content, because that question of definition has been a confusing factor in some of the discussions hitherto. When we talk in general terms about content, we mean the substance of a message. This has been the source of some confusion. The Bill defines “content”, for the purposes of this legislation, in Clause 207 extremely broadly as

“anything communicated by means of an internet service”.

Under this definition, in essence, all user communication and activity, including recommendations by an algorithm, interactions in the metaverse, live streams, and so on, is facilitated by “content”. So, for example, unwanted and inappropriate contact from an adult to a child would be treated by the Bill as content harm. The distinctions that the four Cs make between content, conduct and contact risks is therefore not necessary. For the purposes of the Bill, they are all content risks.

Secondly, I know that there have been concerns about whether the specific risks highlighted in the new schedule will be addressed by the Bill.

Where are the commercial harms? I cannot totally get my head around my noble friend’s definition of content. I can sort of understand how it extends to conduct and contact, but it does not sound as though it could extend to the algorithm itself that is driving the addictive behaviour that most of us are most worried about.

In that vein, will the noble Lord clarify whether that definition of content does not include paid-for content?

I was about to list the four Cs briefly in order, which will bring me on to commercial or contract risk. Perhaps I may do that and return to those points.

I know that there have been concerns about whether the specific risks highlighted in the new schedule will be addressed by the Bill. In terms of the four Cs category of content risks, there are specific duties for providers to protect children from illegal content, such as content that intentionally assists suicide, as well as content that is harmful to children, such as pornography. Regarding conduct risks, the child safety duties cover harmful conduct or activity such as online bullying or abuse and, under the illegal content safety duties, offences relating to harassment, stalking and inciting violence.

With regard to commercial or contract risks, providers specifically have to assess the risks to children from the design and operation of their service, including their business model and governance under the illegal content and child safety duties. In relation to contact risks, as part of the child safety risk assessment, providers will need specifically to assess contact risks of functionalities that enable adults to search for and contact other users, including children, in a way that was set out by my noble friend Lord Bethell. This will protect children from harms such as harassment and abuse, and, under the illegal content safety duties, all forms of child sexual exploitation and abuse, including grooming.

I agree that content, although unfathomable to the outside world, is defined as the Minister says. However, does that mean that when we see that

“primary priority content harmful to children”

will be put in regulations by the Secretary of State under Clause 54(2)—ditto Clause 54(3) and (4)—we will see those contact risks, conduct risks and commercial risks listed as primary priority, priority and non-designated harms?

I do not want to make my speech twice, but in my final sentence I said that my challenge to the Government is to have a very simple way forward by other means, if those things were articulated, but my understanding is that they are to bring forward content harms that describe only content as we normally believe it.

I have tried to outline the Bill’s definition of content, which I think will give some reassurance that other concerns that noble Lords have raised are covered. I will turn in a moment to address priority and primary priority content, if the noble Baroness will allow me to do that, and then perhaps intervene again if I have not done so to her satisfaction. I want to set that out and try to keep track of all the questions which have been posed as I do so.

For now, I know there have been concerns from some noble Lords that if functionalities are not labelled as harm in the legislation they would not be addressed by providers, and I reassure your Lordships’ House that this is not the case. There is an important distinction between content and other risk factors such as, for instance, an algorithm, which without content cannot risk causing harm to a child. That is why functionalities are not covered by the categories of primary, priority and priority content which is harmful to children. The Bill sets out a comprehensive risk assessment process which will cover content or activity that poses a risk of harm to children and other factors, such as functionality, which may increase the risk of harm. As such, the existing children’s risk assessment criteria already cover many of the changes proposed in this amendment. For example, the duties already require service providers to assess the risk of harm to children from their business model and governance. They also require providers to consider how a comprehensive range of functionalities affect risk, how the service is used and how the use of algorithms could increase the risks to children.

Turning to the examples of harmful content set out in the proposed new schedule, I am happy to reassure the noble Baroness and other noble Lords that the Government’s proposed list of primary, priority and priority content covers a significant amount of this content. In her opening speech she asked about cumulative harm—that is, content sent many times or content which is harmful due to the manner of its dissemination. We will look at that in detail on the next group as well, but I will respond to the points she made earlier now. The definition of harm in the Bill under Clause 205 makes it clear that physical or psychological harm may arise from the fact or manner of dissemination of the content, not just the nature of the content—content which is not harmful per se, but which if sent to a child many times, for example by an algorithm, would meet the Bill’s threshold for content that is harmful to children. Companies will have to consider this as a fundamental part of their risk assessment, including, for example, how the dissemination of content via algorithmic recommendations may increase the risk of harm, and they will need to put in place proportionate and age-appropriate measures to manage and mitigate the risks they identify. I followed the exchanges between the noble Baronesses, Lady Kidron and Lady Fox, and I make it clear that the approach set out by the Bill will mean that companies cannot avoid tackling the kind of awful content which Molly Russell saw and the harmful algorithms which pushed that content relentlessly at her.

This point on cumulative harm was picked up by my noble friend Lord Bethell. The Bill will address cumulative risk where it is the result of a combination of high-risk functionality, such as live streaming, or rewards in service by way of payment or non-financial reward. This will initially be identified through Ofcom’s sector risk assessments, and Ofcom’s risk profiles and risk assessment guidance will reflect where a combination of risk in functionalities such as these can drive up the risk of harm to children. Service providers will have to take Ofcom’s risk profiles into account in their own risk assessments for content which is illegal or harmful to children. The actions that companies will be required to take under their risk assessment duties in the Bill and the safety measures they will be required to put in place to manage the services risk will consider this bigger-picture risk profile.

The amendments of the noble Baroness, Lady Kidron, would remove references to primary priority and priority harmful content to children from the child risk assessment duties, which we fear would undermine the effectiveness of the child safety duties as currently drafted. That includes the duty for user-to-user providers to prevent children encountering primary priority harms, such as pornography and content that promotes self-harm or suicide, as well as the duty to put in place age-appropriate measures to protect children from other harmful content and activity. As a result, we fear these amendments could remove the requirement for an age-appropriate approach to protecting children online and make the requirement to prevent children accessing primary priority content less clear.

The noble Baroness, Lady Kidron, asked in her opening remarks about emerging harms, which she was right to do. As noble Lords know, the Bill has been designed to respond as rapidly as possible to new and emerging harms. First, the primary priority and priority list of content can be updated by the Secretary of State. Secondly, it is important to remember the function of non-designated content that is harmful to children in the Bill—that is content that meets the threshold of harmful content to children but is not on the lists designated by the Government. Companies are required to understand and identify this kind of content and, crucially, report it to Ofcom. Thirdly, this will inform the actions of Ofcom itself in its review and report duties under Clause 56, where it is required to review the incidence of harmful content and the severity of harm experienced by children as a result of it. This is not limited to content that the Government have listed as being harmful, as it is intended to capture new and emerging harms. Ofcom will be required to report back to the Government with recommendations on changes to the primary priority and priority content lists.

I turn to the points that the noble Lord, Lord Knight of Weymouth, helpfully raised earlier about things that are in the amendments but not explicitly mentioned in the Bill. As he knows, the Bill has been designed to be tech-neutral, so that it is future-proof. That is why there is no explicit reference to the metaverse or virtual or augmented reality. However, the Bill will apply to service providers that enable users to share content online or interact with each other, as well as search services. That includes a broad range of services such as websites, applications, social media sites, video games and virtual reality spaces such as the metaverse; those are all captured. Any service that allows users to interact, as the metaverse does, will need to conduct a children’s access assessment and comply with the child safety duties if it is likely to be accessed by children.

Amendment 123 from the noble Baroness, Lady Kidron, seeks to amend Clause 48 to require Ofcom to create guidance for Part 3 service providers on this new schedule. For the reasons I have just set out, we do not think it would be workable to require Ofcom to produce guidance on this proposed schedule. For example, the duty requires Ofcom to provide guidance on the content, whereas the proposed schedule includes examples of risky functionality, such as the frequency and volume of recommendations.

I stress again that we are sympathetic to the aim of all these amendments. As I have set out, though, our analysis leads us to believe that the four Cs framework is simply not compatible with the existing architecture of the Bill. Fundamental concepts such as risk, harm and content would need to be reconsidered in the light of it, and that would inevitably have a knock-on effect for a large number of clauses and timing. The Bill has benefited from considerable scrutiny—pre-legislative and in many discussions over many years. The noble Baroness, Lady Kidron, has been a key part of that and of improving the Bill. The task is simply unfeasible at this stage in the progress of the Bill through Parliament and risks delaying it, as well as significantly slowing down Ofcom’s implementation of the child safety duties. We do not think that this slowing down is a risk worth taking, because we believe the Bill already achieves what is sought by these amendments.

Even so, I say to the Committee that we have listened to the noble Baroness, Lady Kidron, and others and have worked to identify changes which would further address these concerns. My noble friend Lady Harding posed a clear question: if not this, what would the Government do instead? I am pleased to say that, as a result of the discussions we have had, the Government have decided to make a significant change to the Bill. We will now place the categories of primary priority and priority content which is harmful to children on the face of the Bill, rather than leaving them to be designated in secondary legislation, so Parliament will have its say on them.

We hope that this change will reassure your Lordships that protecting children from the most harmful content is indeed the priority for the Bill. That change will be made on Report. We will continue to work closely with the noble Baroness, Lady Kidron, my noble friends and others, but I am not able to accept the amendments in the group before us today. With that, I hope that she will be willing to withdraw.

I thank all the speakers. There were some magnificent speeches and I do not really want to pick out any particular ones, but I cannot help but say that the right reverend Prelate described the world without the four Cs. For me, that is what everybody in the Box and on the Front Bench should go and listen to.

I am grateful and pleased that the Minister has said that the Government are moving in this direction. I am very grateful for that but there are a couple of things that I have to come back on. First, I have swiftly read Amendment 205’s definition of harm and I do not think it says that you do not have to reach a barrier of harm; dissemination is quite enough. There is always the problem of what the end result of the harm is. The thing that the Government are not listening to is the relationship between the risk assessment and the harm. It is about making sure that we are clear that it is the functionality that can cause harm. I think we will come back to this at another point, but that is what I beg them to listen to. Secondly, I am not entirely sure that it is correct to say that the four Cs mean that you cannot have primary priority, priority and so on. That could be within the schedule of content, so those two things are not actually mutually exclusive. I would be very happy to have a think about that.

What was not addressed in the Minister’s answer was the point made by the noble Lord, Lord Allan of Hallam, in supporting the proposal that we should have in the schedule: “This is what you’ve got to do; this is what you’ve got to look at; this is what we’re expecting of you; and this is what Parliament has delivered”. That is immensely important, and I was so grateful to the noble Lord, Lord Stevenson, for putting his marker down on this set of amendments. I am absolutely committed to working alongside him and to finding ways around this, but we need to find a way of stating it.

Ironically, that is my answer to both the noble Baronesses, Lady Ritchie and Lady Fox: we should have our arguments here and now, in this Chamber. I do not wish to leave it to the Secretary of State, whom I have great regard for, as it happens, but who knows: I have seen a lot of Secretaries of State. I do not even want to leave it to the Minister, because I have seen a lot of Ministers too—ditto Ofcom, and definitely not the tech sector. So here is the place, and we are the people, to work out the edges of this thing.

Not for the first time, my friend, the noble Baroness, Lady Harding, read out what would have been my answer to the noble Baroness, Lady Ritchie. I have gone round and round, and it is like the Marx brothers’ movie: in the end, harm is defined by subsection (4)(c), but that says that harm will defined by the Secretary of State. It goes around like that through the Bill.

There are three members of the pre-legislative committee in the Chamber. We were very clear about design features, and several members who are not present were even clearer. So I hear where we are with the Bill, but I have been following it for five years and have been saying the same thing, so if we are a little late to the party I do not think that is because of me. I do not want to delay the Bill but I want to stamp the authority of Parliament on the question of how harm happens, as well as what it is.

My last sentence has to be: let us remember our conversation about trying to measure illegal harm and then think about it at scale for children. We have to have something softer than that; we cannot do it for each piece of content. The saving grace of the Bill is its systems and processes: it will make the tsunami a trickle—that is what we want to do. It is not to say that young people should not have access to the internet. Although I spent quite a lot of time disagreeing with the noble Baroness, Lady Fox, today, I absolutely agree with her about evolving capacities, and I hope that we revisit that question later. With that, I beg leave to withdraw my amendment.

Amendment 20 withdrawn.

Amendment 21 not moved.

Amendment 21A

Moved by

21A: Clause 10, page 10, line 1, after “19(2)” insert “and (8A)”

Member’s explanatory statement

This amendment inserts a signpost to the new duty in clause 19 about supplying records of risk assessments to OFCOM.

Amendment 21A agreed.

Clause 10, as amended, agreed.

Clause 11: Safety duties protecting children

Amendment 22 not moved.

Amendment 22A

Moved by

22A: Clause 11, page 10, line 6, at end insert “(as indicated by the headings).”

Member’s explanatory statement

This amendment provides clarification because the new duty to summarise children’s risk assessments in the terms of service (see the amendment in the Minister’s name inserting new subsection (10A) below) is imposed only on providers of Category 1 services.

Amendment 22A agreed.

House resumed.

House adjourned at 6.48 pm.