The Online Safety Act 2023, which recently received Royal Assent, has been designed to keep pace with emerging technologies. The Act will regulate AI-generated content in much the same way that it does content created by humans. It covers AI-generated content shared by users with other users, search results generated by AI and AI-generated pornography. In addition, the Act will criminalise the sharing of deepfake intimate image abuse, including when that is AI-generated.
May I take this opportunity to welcome the Minister to the Dispatch Box? I will not embarrass him by calling him a friend, but I wish him the best of luck.
As the world transitions into the new age of AI, the Tories are leaving the people of Scotland at risk from harmful AI-generated content and social media. Will the Minister outline why the Scottish Government was blocked from participating in one of the first major AI safety summits?
I thank the hon. Member—I will embarrass her and say that we are friends. However, I respectfully disagree, because the Scottish people were represented by the UK Government.
Meta, having recognised the threat that unrestricted use of AI could represent in elections and democracy, has banned the use of generative AI in its political adverts. Why does a private company seem to be doing more to curb the spread of electoral misinformation than this Government?
The Government take the integrity and security of our democratic processes very seriously. We will continue to safeguard against future risks, strengthen our resilience and ensure that the regulatory framework is as effective as possible. DSIT supports wider cross-Government efforts to protect UK democratic processes, including through the defending democracy taskforce and election cell, and will be working closely with social media platforms to ensure that the right systems are in place to identify and remove harmful material, including deepfakes, where it breaches platform terms of service.
I call the Chair of the Select Committee.
I warmly congratulate my hon. Friend on his appointment and welcome him to the Dispatch Box.
GCHQ this week says that it expects hostile disruption of the next election through deepfakes using AI. Is my hon. Friend absolutely confident that the Electoral Commission has all the powers it needs to prevent that, and why are the Government not implementing their suggestion in the White Paper to introduce the legislation to empower regulators?
I thank my right hon. Friend for his kind remarks. As I said, the Government will take this issue incredibly seriously, and I am confident that, through the defending democracy taskforce and election cell, we will be able to do the utmost to protect ourselves from election interference. I offer to meet him to discuss this further and see what else can be done.
I too welcome my hon. Friend to his role. What discussions has he or his Department had about raising public awareness of the impact that AI will have on society, and increasing understanding that not all content, harmful or otherwise, might be what it seems?
I thank my hon. Friend for his remarks. We have just had a global, leading AI safety summit, which had immense coverage on that and also focused on frontier risks. We have always been clear that we will take a pragmatic, proportionate and contextual approach. With the 28 countries plus the European Union who have agreed to the Bletchley declaration, there is a great opportunity to use AI for our benefit, but we should also be wary of the risks involved.