It is a rather glorious day, so if people would like to remove their jackets, they should please feel free to do so.

I beg to move, That this House has considered Government policy on AI and its impact on society. It is a pleasure to serve with you in the Chair, Ms McVey. I think we all agree that AI is a transformational technology with the potential to bring many benefits to our society, but to fully realise them we will require the Government to look at radical changes to taxation, welfare and our industrial strategy. Everyone in the UK needs to benefit from a managed AI transition that puts workers and human dignity at its heart. It must make the world fairer, not more unequal, and it should give UK citizens a meaningful say in decisions that will affect their lives. Evidence from the New Contract, a pro-worker AI campaign organisation, reveals that the public are deeply suspicious about AI. Around six in 10 people expect the gains to flow to wealthy investors and big corporations, while just 7% think they will be shared fairly across our society. Understandably, seven in 10 workers are worried about the impact of AI on their jobs. Even the Foreign Secretary said recently that AI poses a Hiroshima-style threat to humanity unless global rules are put in place. In today’s digital age, a cartel of technology giants—Google, Meta, Microsoft, Amazon and Apple—wield extraordinarily influence over our lives online. These platforms have become so embedded in daily life that meaningful participation in society depends on using their services. Similarly, the digital infrastructure that facilitates our work and public services is now an essential layer of the economy, but that gives oligarchs like Elon Musk enormous power to distort public discourse. We have effectively subcontracted our right to information to a handful of big tech gatekeepers.

The hon. Gentleman is making a powerful point about big foreign tech companies. Does he believe that our dependency on six tech companies in all aspects of our lives poses a national security threat and needs to be assessed as such?

I thank the hon. Member for his intervention; I will deal with AI sovereignty a bit later in my speech. We should ensure that, like a socialist Government’s approach to energy or water, the public have a stake in the development of AI technology, to ensure that the value it creates is captured and shared for the good of society as a whole. Previous waves of technological change have brought with them huge economic disruptions, but the human and social costs that followed were not inevitable. They were the product of political choices, as Governments left workers and communities to absorb the shocks alone.

The last time we saw a change of such potential scale was when we moved from being a manufacturing economy to being more of a service economy. There were trade-offs at that time, and some people, cities, towns and villages were left behind—including in my constituency. Does my hon. Friend agree that on this occasion we can plan because we know it is coming? We need to be thinking about jobs, access and skills for younger people, and ensuring that the benefits are felt right the way across the country and not just clustered in one place.

I absolutely agree with my hon. Friend’s point about no place being left behind as a result of the AI revolution. I will develop that point a later. We can already see the first signs of AI impacting on work here. We have record numbers of young people not in employment, education or training because many entry-level jobs have disappeared.

I congratulate the hon. Gentleman on securing the debate. I spoke to him beforehand; he knows where I am coming from. My concern is for the young people of today looking for the jobs of tomorrow. My grandchildren will have to prepare themselves. We must try to equip young people. In the last few days the Belfast Telegraph reported that this may be the first time in a century that young people have lower scores in tests for memory, focus and reading than their parents. Is it not time for the Government to acknowledge the challenges and ensure that there are employment prospects for young people? If the hon. Gentleman’s objective is that, it is the right way to go.

Yes, young people and their employment prospects are intrinsically linked to the debate about the future of AI. The hon. Gentleman is right that we have to get that right for those people to have a future. We also need the Government to introduce a raft of measures to manage AI-driven changes to work. Those measures should include an employment levy on companies that replace large numbers of workers with AI, training subsidies for displaced workers, redeployment programmes to move workers into sectors with skills shortages, and a job subsidy scheme for workers and companies at the sharp end of industrial change, along with, of course, stronger social security support for those who face unemployment. We know that if it is left to the market, firms will often reach for the crudest form of automation, stripping out roles and degrading the work that remains. The alternative is that we start to advance the case for dignified work, and recognise that that will mean having meaningful worker involvement in every step of the process. The involvement of workers in adoption is often what turns a promising tool into a productive one. That is why I agree with the TUC that the Government should strengthen the Employment Rights Act 2025 to give workers a real say in decisions. There should be a duty on employers to disclose their use of AI, and a right for the workforce to be consulted and negotiate over how AI is introduced. One of the biggest challenges of AI is to ensure that everyone gets a fair share of the gains. The AI and big tech corporations in the US stand shortly to become some of the wealthiest and most powerful private actors in history. Their wealth has been built on public research and public investment, as well as all our shared written inheritance, which has trained the models they use. The fact that our tax system taxes income from work far more heavily than income from capital gives firms a direct tax incentive to automate a worker out of a job rather than employing them, and could see AI facilitate a further shift of national income away from labour and towards capital, eroding the tax base that funds the public services that we all need and want.

The hon. Gentleman is focusing on employment and big tech companies; does he agree that we need to go a step back—back to school, teachers and training? Teachers need to prepare their pupils for the future with AI, and for what that means. Does the hon. Gentleman feel that that is something not for this Minister but perhaps for a Minister in another Department?

I thank the hon. Gentleman for his second intervention—let us see whether he can make it a hat trick before I finish. The reality is that AI is here now. We have to deal with how it impacts on today’s workers and future workers as well. This is not one generation against another, and I know the hon. Gentleman understands that. The Government could seek to equalise tax treatment, actively explore widening the digital services tax to include AI companies, which it currently excludes, and raise the level of the tax from 2% to the European average. We cannot allow the use of machines to be made artificially cheaper than employing people. It is important to look at AI sovereignty, which the hon. Member for Dewsbury and Batley (Iqbal Mohamed) mentioned earlier. The UK is growing increasingly dependent on US tech, and owns little of the data infrastructure and AI models on which the future economy will increasingly rely. This repeats the pattern of recent decades in other sectors, whereby essential national infrastructure has passed into overseas ownership and money has flowed offshore, leaving British people with less and less control over the essentials of a good life. That dependence not only creates security risks, as has been mentioned, but prevents the UK from capturing and distributing any AI windfall. We need to give UK companies a meaningful chance to compete. The Competition and Markets Authority could use existing powers to take on the concentration in the cloud market, which underpins the AI companies’ business models. A progressive sovereign AI programme should involve an industrial strategy that seeks to expand public investment in the sector and impose strict conditions on private access to public assets, alongside active support for models such as co-operatives, public interest companies and other democratic ownership models, so that the gains of AI can be captured and shared throughout the UK. At the heart of this debate is the role of the public. Big decisions about AI, such as where data centres are built and how AI is deployed in public services, are being taken with little democratic input.

I apologise in advance, Ms McVey, that I need to go to a Select Committee so cannot attend the full debate. My hon. Friend is, as usual, right about everything. Does he agree that public resistance to the construction of data centres is due not just to concern about the impact on the environment, but to underlying public scepticism about the breathlessness with which politicians talk about economic growth driven by digital technology when we have an economy that has a completely unsustainable food system, is not delivering genuinely affordable homes, has rubbish public transport in many parts of the country, and is not delivering the basics?

One data centre, even a modest one, uses the same amount of electricity as 100,000 homes. The bigger the data centre, the more electricity it uses. Very few of them run off renewables, so that is another issue we need to address. The communities that host the infrastructure see the costs, in water, energy and land, without getting any of the benefits that we have discussed. These big debates are why the Government must embed ongoing public and democratic oversight of the AI transition and legislate for community benefit where infrastructure is built, including a share of the value created locally. We must ensure that the AI revolution does not lead to more power being in the hands of a few who can determine our future. Governments must build independent, publicly funded alternatives to ensure that AI is developed for the common good. Only public investment can support AI that prioritises social and environmental challenges. We need an “AI for the people” strategy—one that starts to recognise the challenges we face and has the ideas to meet those challenges head on.

Order. I will call the Front Benchers from 3.30; I will not impose a fixed time limit, but about six or seven people want to speak.

I rise in support of much of what the hon. Member for Poole (Neil Duncan-Jordan) has said. He has done a service to this Chamber, and Parliament more generally, by highlighting the threats that AI will bring. Perhaps I can set the scene by reminding Members present that it took 25 years for Parliament and successive Governments to recognise that the internet must be regulated. That is extraordinary. When I was in the IT industry back in the ’80s and ’90s, I became aware of the concept of the internet—before it was widely spoken about—and even then I knew the harm it might do. In 2000, at the beginning of the new millennium, I was privileged to be able to speak at the Spalding Gentlemen’s Society, with which I do not imagine you are familiar, Ms McVey, but suffice to say it is a learned society of the old school—it now does admit women, you will be pleased to know. It has a lecture programme on all manner of things, and I was pleased to be invited to give the first lecture of the millennium. Then, I warned of what the internet might do. Most of the audience, by the way, disbelieved what I said. It was a very civilised meeting—it was not raucous in any way—but they could not quite come to terms with what I predicted. What I predicted then has come to pass. What has occurred is the distortion of public discourse, the enabling of widespread fraud and the provision of the means for every imaginable form of harm, from hardcore pornography being available to children, to internet gambling addiction, suicide sites and much more. Now, the prospect of AI will make all that still more chilling. As the recognition of what is real and what is invented is increasingly blurred, it will become more and more difficult for people to navigate the inevitable vicissitudes that pervade every life. Imagine a future where the once secure certainty of what really exists is so undermined that people can no longer believe what they are told and have no place to go to establish that certainty—the basis on which we all gauge reality and are able to live reasonably productive lives. I see wellbeing being affected by that detrimentally; unless the Government take very urgent and extremely serious action, that will happen far more quickly than anyone here envisages.

Does the right hon. Gentleman agree that one of the primary obligations of any Government is to keep their citizens and the people living in their country safe from all avoidable harms? AI is a major potential cause of harm, and it is avoidable. Does he agree that the Government cannot wait 25 years to catch up?

I absolutely agree. That is why I began by describing what has occurred over the last quarter of a century and the misunderstanding—a pervasive one that affected all political parties and Governments of all kinds—that a dangerous fascination with novelty was more important than the certainty that I have just outlined. The hon. Member for Poole is right to issue a clarion call for action. Let me go into a little more detail. He is right, for example, that we have allowed a handful of multinational corporates, which have displayed a carelessness to the effects of what they do, to wield immense power—more power than national Governments. It is increasingly difficult for Governments to pull levers that affect these people, because they exist beyond the purview of this House or of any Government Minister. They have shown by their example that carelessness, and only when they have been prevented from doing harm have they ceased to do so. I am delighted that the Government are now taking steps to prevent children from accessing social media and mobile telephones; I would like to go much further, actually, and I hope the Government will. I called for the ban on phones in schools, along with many other Members of this House, a long time ago. But AI makes all that more serious, because the AI available through that technology, which has the ability to distort the very nature of reality, will be immensely difficult to control. And yet we must try to do so. Legislation is not always the only means of doing that but, as the hon. Gentleman set out, it certainly must play a part. Umberto Eco described the internet as the “empire of imbeciles”. Over time, it has become impossible to distinguish experts from idiots, and every bar-room bigot now has an audience of millions. Most people had a sense of proportion; my parents were working-class people who left school at 14, but they had a very good sense of what really mattered. Now we have elevated trivia to an extraordinary degree and, as a result, that sense of proportion, which was once taken as read, can no longer be guaranteed.

On the point about proportionality, a lot of concerns about AI have rightly been raised in the debate, but let me sound a more positive note about public services. In my constituency, North Kerrier integrated neighbourhood team, which covers one of the most deprived regions in the UK, is using Brave AI to identify frailer and more vulnerable patients earlier and thereby prevent avoidable hospital admissions. Given the opportunities of AI, and the risks the right hon. Gentleman is talking about in great detail, does he agree, particularly in relation to healthcare, that the right approach is proportionate regulation, so that we do not stifle innovation, but maintain public trust and safety?

I know from my dealings with the hon. Gentleman that he is a good and effective Member of Parliament who takes a very responsible view of these things, but we must not let the virtues that he describes blind us to the vices. That is what happened with the internet. When I made that speech back in 2000, when people said, “But I can reunite with my family in Australia.” I replied, “Well, if they really wanted to unite with you, they wouldn’t have gone to Australia in the first place.” I remember Friends Reunited—as though we want to get in touch with people we were at primary school with. They look so different; that is frightening in itself. Why would we want to revisit our ancient history? But that is all an aside. I think the hon. Gentleman is right, and of course it is true that there will be the advantages that he describes, particularly in the research field. My fear is that they will obscure, in the eyes of those who have the power to make a difference, the points that the hon. Member for Poole made. I hope that does not apply to this Minister, who I am sure is far more clear-sighted than that suggests. Parliament and the Government need to act quickly and decisively to deal with the points that the hon. Member for Poole made and that I have attempted, imperfectly, to amplify. It is critically important to understand that although innovation may matter, ethics matter more; that science and technology are morally neutral and have no implicit ethic; and that what really counts is not what is new, but what is true.

It is a real pleasure to speak in this debate. I am very grateful to my hon. Friend the Member for Poole (Neil Duncan-Jordan) for securing it, and I agree with many of his remarks and those made by other speakers. I particularly agree with his remarks about the New Contract initiative. The UK’s AI sector has grown 23 times faster than the rest of our economy. We must ensure continued investment and the realisation of productivity gains from appropriate use, but we must also secure greater domestic control and resilience, and mitigate near and later-term risks from inappropriate use, if we are truly to become an innovation nation. I believe that we must change our approach to AI, and I want to spend the time that I have setting out the case for doing that. Above all, we must seek sovereignty as if we actually mean it. Supporting UK niches, such as photonics and quantum, is key to sovereignty and to boosting growth from Southampton to Sheffield and from Belfast to Birmingham, but we must recognise how crucial access to frontier models is to our sovereignty. The recent debacle with Fable 5 and Mythos 5 only serves to underline that. The sovereign AI fund needs a clearer rationale. It should be developing UK companies and partnerships, and public stakes should genuinely put our people and our places first. We must also work more concertedly with allies to develop joint sovereign capability. AI development needs to be part of an economic security reset with the EU, and we need to be able to participate in the Scaleup Europe Fund. Interestingly, Mistral AI’s chief executive recently claimed that the EU has at most two years to build its own AI or become permanently structurally dependent on the US or China. The UK must be included in that EU-level debate before it is too late. Delivering stronger sovereign control and resilience also means the Government putting their money where their mouths are and supporting UK AI and UK partnership AI in procurement. All Departments should be assessed publicly on that, and the Competition and Markets Authority should be engaged on how to ensure more advanced market commitments to support our sovereignty. We can learn from other countries on that. Look at France, which recently dropped Palantir from its intelligence services. Of course I realise that changes cause disruption, but we cannot work purely on the basis of cost or short-term capability any more. We have to make assessments on the basis of control and resilience, given their significance for our national interest. In public services, we must move beyond no, low and performative adoption to appropriate and transparent adoption that puts people and places first. It is critical that the UK can deploy sufficient compute to support those aspirations. I hope that the AI Energy Council can develop plans for a line delivery of renewable energy and data centre projects, as well as fixing difficult issues such as demand modulation to unblock additional capacity. There can genuinely be a win-win here, but for that and greater adoption to be achieved, I agree with what others have said: we have to face up to public concern about the risks. We need only look at what is happening in Scotland right now to see that. First, on unemployment, half of all UK workers feel that they will lose their job to AI. From what I can see, there is very mixed evidence about that so far, beyond the field of software development. Pleased as I was to see the Early Careers Jobs Alliance and the AI adoption insights agreement from the Government, we must go further. Lessons should be learned from the consultation on the social media ban, “Growing up in the online world: a national consultation”. I would like to see the Government launch a consultation on starting work in a world of AI, so that we can involve young people who are worried about that, gather their good ideas and show that the Government are engaging with that crucial issue. We should be doing far more in our schools and colleges to prepare our children.

AI is increasingly being used to make cutting-edge assistive technology for people with special educational needs and disabilities, but the industry is a bit of a wild west, and access for young people in areas such as Yeovil is uneven. Will the hon. Member join me in urging the Government to publish guidance for schools on best practices for AI assistive technology for SEND?

The hon. Member raises a really important point, and it is part of a broader picture. My understanding is that something like only half of all schools have an appropriate AI policy. Simply saying to students, “Don’t use AI”, having an unorganised approach to it, or just taking a company’s word for it without a broader set of guidance and frameworks for deployment is not supporting our children and young people. Saying, “Just don’t use AI”, when we know children and young people are using it, is not fair to them, and will do a disservice to our economy in the future. Our children and young people need to understand how AI works, as well as the risks of cognitive offloading and the negative impacts on social interaction from inappropriate use. I also want to see us encouraging international discussion in the OECD and the United Nations about taxation—taxation of permanent establishment, capital gains and the cost of human labour compared with tokens. That does not need to have an immediate impact; we could set a threshold of employment impact to trigger a token tax, for example. However, we cannot run away from this debate, especially given how much services dominate our economy. Secondly, we have to engage with concerns about the safety of frontier models. As I said many months ago alongside the Centre for Countering Digital Hate, just regulating applications after harm has been done is unacceptable. Grok should not have been able to produce 3 million non-consensual nudified images in the first place. New York, California, Illinois and the EU have moved towards approaches requiring mandatory assessment of frontier models, and we should too. We should place the AI Security Institute on a statutory footing as a first step, and we should use our presidency of the G20 next year to reinvigorate global attempts at co-ordination, which started with the Bletchley conference but have effectively stalled since. We can work with EU countries to do that as well as with nations from India to Brazil. Thirdly, and finally, although AI can release huge productivity gains, we have to recognise that unscrupulous and inappropriate use is already imposing huge costs too.

I am interested in the right hon. Lady’s comment on productivity, because it is one of the prevailing macroeconomic challenges facing this Government and previous Governments, as the stalling of productivity has been a problem across western economies for a while. It is true that AI might help, but it could also harm. Equipping people with skills is critical to building productivity, and dependence on AI may actually undermine people’s confidence and their skills over time.

I absolutely agree. That phenomenon of so-called cognitive offloading and the lack of scaffolding in skills is a genuine concern. Equally, we see how rapidly and accurately AI can assess tumour scans, for example, working with human beings to deliver the highest possible quality of healthcare. I agree with my hon. Friend the Member for Camborne and Redruth (Perran Moon) about the potential there. We need to have a sensible approach to this issue. I also want to mention some additional harms. We are now seeing ubiquitous slop on the internet, AI-enabled fraud and cyber-attacks, and the impact of AI summaries on local media use as well, so I want to see the CMA using its powers for competition-focused action on strategic market status services. I want the Government to adopt a maximalist approach to the “Watch this Space” White Paper, and I want us to stop treating bots as separate, because AI is increasingly going to be integral to our interface with the internet, as well as to so much else, so we need to adopt a safety-by-design approach to online regulation, and that needs to recognise the increasing ubiquity of AI.

I commend the hon. Member for Poole (Neil Duncan-Jordan), as well as his two friends and allies who have just spoken, my right hon. Friend the Member for South Holland and The Deepings (Sir John Hayes) and the right hon. Member for Oxford East (Anneliese Dodds), on bringing this important subject to Westminster Hall. This is in danger of turning into one of those rare debates in Parliament where one learns something from everybody who speaks and we all come away better informed and equipped. This subject is enormous, and it is only a 90-minute debate. It obviously will not be the last time we discuss these matters, and I freely confess to not being an expert on the subject. Like many colleagues, I am trying to educate myself more to understand better the implications, including the upside for productivity, the implications for energy and water, and the effect on all manner of things in the public and private spheres, such as healthcare, education, international relations, and foreign and security policy and so on. I am no expert, but one thing I have learned is that no one knows what the societal effect of AI is at this moment in time. A lot of people have a story to tell, but no one really knows. That is one of the reasons why we have to keep coming back to these subjects as our understanding iterates and evolves.

The right hon. Gentleman says no one knows, but throughout history, when we have approached technologies or situations that we have never experienced before, we have done scenario planning and risk assessments, and we have prepared as best as possible for scenarios that may be plausible or possible. Does he agree that when it comes to AI, we have been absolutely asleep at the wheel and that planning has not happened?

I agree with large parts of what the hon. Gentleman says. The timeframes are obviously shorter this time than they were for say, the industrial revolution, the printing press or any of the other massive changes that we have seen in the past. When Gutenberg invented his printing press, I am not sure it was foreseeable to even the most far-sighted, omniscient individual what all the effects of that would be. That does not mean we do not try to scenario-plan and to think about the different possible outcomes and try to mitigate them. Labour market effects are probably what we as politicians worry about the most, along with all the other implications, like the consequences for income distribution, intergenerational fairness and so many other aspects. People already say that AI has had a big impact on job markets, particularly on the graduate job market. I honestly do not know whether that is true or not. It strikes me that it is quite an easy thing to say: if a company is not doing that well this year and is not employing graduates to the same extent that it used to, it is a great thing to say, “Well, we’re investing in AI.” I do not see how a professional services firm, with the current technology deployment, would be able to substitute people for AI in quite that way. In the future, it is true that there is a scenario in which there is mass unemployment. But it is also true that when there have been huge upheavals previously, labour markets have adapted, albeit over time, and sometimes it is that gap that makes the difference. Labour has found its way into different sectors—

I thank the right hon. Gentleman for his excellent speech. There are lots of things about AI and its impact that we do not know. But one of the things that history teaches us is that if vast amounts of power and wealth are handed to a handful of individuals, corporations or even the Chinese Communist party—this is not just about corporations; it can be about states as well—the outcome is usually not good for the vast majority of people. I think we can all agree on that. We do not need to know whether AI will be beneficial or hostile, or become super-AI. What we can see is that when a small group of people are allowed to define what reality and knowledge are, that in itself—well, we have all read “1984”.

I agree with the hon. Gentleman; this is, of course, not a left versus right argument. For example, the antitrust regime in the United States exists precisely to prevent the concentration of power. Indeed, all the companies in the list we heard earlier were American. We did not mention ByteDance or Huawei, which we could easily have done. As the right hon. Member for Oxford East said, sovereign capability is obviously important, too. On labour markets, there is also a risk—actually, the hon. Member for Poole did not do this, but I thought that he was going to keep saying, “Universal basic income, universal basic income.” He did not, but a lot of people do. Universal basic income is a policy prescription that those people calling for it tended to believe in years earlier, before they had even heard of AI, and now it fits this scenario. I do not know what the outcomes and therefore the mitigations of AI will need to be, but we must keep an open mind.

Will my right hon. Friend give way?

I had better press on, because we have a number of colleagues to get through. Regarding generative AI and its effect on information integrity, which my right hon. Friend the Member for South Holland and The Deepings referred to earlier, we should bear in mind that so far the massive growth in fraud, misinformation and disinformation has all been done with cheapfakes. People talk about deepfakes, but it is done with rubbish, which can still take in large numbers of people. Just imagine what happens when such fakes become convincing. For example, there is the email that says, “I am stranded in Bermuda”, or whatever it is, “and I need you to send me £100 overnight.” Imagine when that scam actually involves a video of someone’s child saying it to them. However, we soon realise that those things seem relatively benign when people start talking about agentic AI and super-intelligence, whereby machines take control of a situation, which some people have described as being ultimately an existential threat to humanity. I do not know what all the answers are, but I think that as politicians—as the House of Commons—we need to ensure that our organisational architecture is equipped to consider all these questions as they evolve. However, I do not think that our Select Committee structure today is equipped in that way. For example, I do not think that having a Science, Intelligence and Technology Committee, which obviously covers technology, is a suitable way to address all these matters, and I also do not think that it is appropriate to expect every other Select Committee to develop such expertise themselves. Somehow, we need to meld the two structures. We must have some form of super-Select Committee—a support team, or something like that—to work with all the other teams or Committees across the piece. I will finish by mentioning two specific things that we can do right now, because there is a ticking bomb with AI, specifically regarding children. In education now, if someone asks the typical year 9 pupil, “How many of your friends use AI to help them do their homework?”, that child would just laugh and say, “All of them.” The idea that we have some sort of issue where we have to encourage children to use AI is just nonsense. It is also crazy to suppose that we can just say to them, “Don’t do it,” and they will comply. Every child in the country learns computing in primary school. We need to make sure that the curriculum evolves so that children get to understand how the thing works, how it generates its content, why it might hallucinate, why it tries to flatter the questioner and why we cannot fully rely on it. There is a real danger. Cognitive offloading is a fancy term, but it basically means relying on a machine to do the work for us. One way we can mitigate that effect with schoolchildren is ensure that when they do their exams at the end of the year, they do them with a piece of paper and a pen in exam conditions. I am afraid there is a move across exam boards to make more exams digital. Of course, they will say that the computers are not attached to the internet and children cannot do this, that and the other, but pen and paper is the safest thing. Of course, there should be exceptions for children with a special educational need or disability that means they need to use a keyboard, and exceptions for computer science, but in general, children at school doing their public exams should do them on pen and paper. That will make sure it happens lower down the school as well. It will also ensure that children continue to write. This sounds like a stupid thing to say, but people already ask, “Why is my child learning to write with a pen, because by the time they grow up nobody will be doing that?” Finally, we are about to make the same mistake with chatbots as we did with social media. I am talking about anthropomorphism, where a computer program develops a personality of its own and develops relationships with people. I welcome the development of an age-appropriate design code for chatbots, but it is not enough. The regulatory regime we have in this country, with Ofcom and the Information Commissioner’s Office, relies on there being evidence of harm. The evidence will not come before the harm. We need to change that system and adopt a precautionary principle when it comes to protecting children.

It is a pleasure to serve under your chairship, Ms McVey. I have rewritten my speech because I do not want to repeat all the things that have been said. I think we are in agreement, although we may have different perspectives. I want to take us back to 80 years ago, to a country house a few miles from Milton Keynes in my constituency. A handful of mathematicians built a machine that cracked the enigma code, and it gave birth to modern computing. That was Bletchley Park, where Britain proved that it could outthink the world. That inheritance still runs through Milton Keynes in the data centres rising on our edge, the British Standards Institution at Knowlhill—the oldest standards institute in the world, whose kitemark has told people that a product is safe since 1903—and a plethora of AI start-ups that outnumber Oxford or Cambridge. Britain is standing at another one of those moments, as we see with our economy and our military tech coming together and pushing forward either for the good of the nation or for its ultimate damage. The potential benefits of artificial intelligence are enormous, and not only for the private sector. AI is a tool that can pull our public services back from years of managed decline. The technology will allow us to detect diseases earlier and cut NHS waiting lists. It will free nurses and other public servants from back-office grind for work that only people can do. However, every country can see that prize, and the question that decides whether Britain leads or merely buys in is whether we make the technology or just import it. A maker exports its products, services and ideas around the world, as we did in world war two, but the UK is at risk of becoming an AI taker and simply handing over the keys to giant US tech firms that are buying up our best young AI companies. It is still our technology; it is just being bought up, packaged and sold back to us so that we become indebted. As we are talking about the impact on our economy and our society, I want to point to Palantir, which has the contract with the NHS. We know that UK companies could do exactly the same work, because I and many Members present have spoken to them. Even more concerning is that the Ministry of Defence now uses the company. With the activist Government in the US turning off technology and restricting its availability outside the US, that puts at risk not only our NHS but, most importantly, the defence of our realm against Russia, for which others may have more sympathy than we do. There is also a social cost, as people look at a company such as Palantir and say that it does not reflect their values. It is used by US Immigration and Customs Enforcement and by Israel to target Gaza. That has a huge impact on public trust. It is also bad for the UK economy and innovation, because it crowds out UK firms that could have those public contracts. That is just one example of a dependency that is putting us at risk both socially and economically. It does not have to be that way. British AI firms need three things to succeed: customers who will buy their products, the ability to increase supply to meet demand, and a USP that British AI is the most trusted. The Government can help with all three. First, we have to act as a good customer. Why should those NHS and MOD contracts not go to British AI firms? Not only would that increase trust, but those products would be developed in a way that meets our needs; we would not be buying something that had been developed elsewhere. That would also mean that we had to invest in the entire British AI stack—it must be British from start to finish. The second issue is supply. We need to make sure that we have the compute power we need and that we are not relying on US firms again for that. We are overly reliant: despite the fact that we part-own Crown Hosting, that accounts for only 6% of public sector data; the rest is nearly all on Amazon Web Services, which is a significant risk. We need to look at how we diffuse the supply of compute power—the underused capacity—to increase that. Finally, we need to increase AI take-up across British businesses by building trusted products. The British Standards Institution is at the heart of that. It does not always have to be about regulation; it can be about standards that lift everything up. What attracts the world to the BSI is that it is known for doing that and its standards become international, so we get first-market advantage. This partnership model of an industrial strategy is mission-driven, creates public trust, evaluates the impact on society and creates opportunities for our ideas and innovation to flourish. That is something we did with the innovations at Bletchley Park. We need to become an innovation leader again—an AI maker, not just an AI taker, which puts us at risk.

As three Members wish to speak, the time limit for speeches is three to four minutes, because we are going to the Front Benchers at 3.30 pm.

Responsible AI, done right, could be the saviour of humanity. Irresponsible AI, done wrong, could lead to the destruction of humanity. There is, to my knowledge, no effective, self-regulated industry anywhere in the world that puts consumers and the planet before profit and shareholders. I thank the hon. Member for Poole (Neil Duncan-Jordan) for securing this important debate. As he and many Members have highlighted, a handful of companies run by men answerable to no electorate and standing before no constituents are making decisions right now that will shape healthcare, education, employment, childhood and the economy more than most Acts of this Parliament ever will. I will focus on topics that have not been covered and speak about safety and economic risks. I have spent 20-plus years in the pharmaceutical regulatory industry, and we can learn much from it. We regulate medicines before they reach a single patient. We do not wait for harm; instead, we test, we license and we monitor. That is the entire logic of the GxP framework that underpins pharmaceutical safety in this country, yet AI is being deployed in our hospitals, schools, courts and welfare system with no comparable safety regime at all, not because the risk is smaller but because the oversight is simply not there. I put the following demands to the Government. First, where GxP protects patients from medicine, we need GAIP—good AI practice—regulation to protect patients from algorithms. Any AI system used in healthcare must meet a statutory standard before it touches a single patient record or clinical decision. The AI Security Institute already evaluates frontier models for biological, chemical and cyber risk, but only if the AI companies allow it to. That is progress, but progress in security assessments is not the same as regulating delivery in the healthcare industry. This Government brought Palantir into the heart of our NHS; in doing so, they handed a US surveillance contractor that is complicit in genocide the keys to patient data that belongs to the British public, not to a share price on Wall Street. I ask the Minister to guarantee the British public that their NHS data will never be given to or allowed to be stolen by Palantir or any other private company. Secondly, the Government must mandate independent statutory testing for AI before its release, and a yellow card scheme that would be a national register where safety incidents, near misses and harms caused by AI were reported, tracked and acted on exactly as we do for adverse drug reactions. It must not be the manufacturer marking its own homework and hoping that nobody checks; self-certification does not work and must end. Thirdly, every AI system developed in this country must have a kill switch that is triggered automatically the moment a system takes unauthorised action and that any human being can trigger manually at any time without obstruction or AI override. This danger is not only about safety but about who pays the bill as AI unfolds. More than 300,000 private-hire drivers in this country face displacement by robotaxis. I ask the Minister how the Government will fund public services going forward. There are dangers, but there are opportunities from AI. We must try to access them as much as possible, but we must address the societal, economic and safety consequences that AI poses.

It is a pleasure to serve under your chairship, Ms McVey. I thank my hon. Friend the Member for Poole (Neil Duncan-Jordan) for securing this crucial debate. The quality of the contributions reflects how timely this conversation is. Artificial intelligence is no longer a distant future challenge; it is here, shaping our economy, public services and daily lives. The challenge is how we ensure that these changes work for people, strengthen our communities and deliver benefits that are shared across the country and internationally. The UK has an opportunity to lead by embracing innovation and making sure that the right safeguards are in place to protect our communities, build confidence and ensure that the benefits are shared. It is vital that we recognise how we can share in that new economy, and that we act now to safeguard public value and wealth generation. Every day, through our public services, our NHS, our transport network and countless interactions with the British public, we generate data that has enormous value. AI is the technology that unlocks that value, but the resource it is built on is our human data, which belongs to us all. As I outlined in a recent article, we must learn the lessons from the past on this. In the past century, Norway decided to turn the national resource of fossil fuels in the North sea—the original driver of modern industrial development—into lasting public wealth. In the UK, we did not invest the revenues from North sea oil; we spent it to cover day to day costs. We must not make that same mistake twice. Rather than allowing the value created by AI from public data to flow permanently into a few private hands, we must ensure that British public services share in the wealth generated from that value—from the national assets we have created—by creating a national data wealth fund. The Government have already recognised the strategic importance of data through the national data library and the AI opportunities action plan. Those are welcome steps forward, but we must do more to build the institutions that capture this value before it slips through our hands. I fully agree with the comments made by my right hon. Friend the Member for Oxford East (Anneliese Dodds) about the need for real sovereign AI capabilities and urgent collaboration with our European allies. Keeping our communities safe and defending our democracy are essential parts of the AI challenge. I recently met the Minister for AI and Online Safety to discuss the growing impact of AI on children and young people particularly. We need to ensure there are strong safeguards around AI-generated content. If we cannot trust what is shown to us online, and if we cannot distinguish between what is real and what has been artificially created, the consequences can reach far beyond individual users and can undermine trust in our democratic institutions themselves. That brings me to accountability. Companies with influence over what millions of people see should be held accountable and made responsible. Social media is not a village noticeboard. It is driven by algorithms that control what is recommended and amplified to millions of people every day. Technology companies act as editors: they decide what people see, what spreads, what is promoted and what disappears. AI will increasingly determine the information that people consume not only through social media, but through search engines, digital assistants and the services people use every day. That is why transparency, accountability and effective regulation are crucial. As AI use grows, so too does the demand for further infrastructure, as I have outlined before. Data centres will become increasingly important to Britain’s economy. As the Member for South East Cornwall, I know that nature is not a blocker to growth. With careful planning, meaningful engagement and strong environmental protections, we can protect the landscapes that make Cornwall special while creating skilled jobs, attracting investment and ensuring that local communities benefit from a new technology, which, concentrated in companies in London as it is, can often feel very far from South East Cornwall. The economic revolution is happening whether we welcome it or not, but our responsibility is to ensure that we are ready to unlock the opportunities, that we protect children and vulnerable people, and that we hold the powerful platforms to account. I am ready to ensure that the remarkable technology being developed today strengthens my community and does not leave them behind.

It is a pleasure to serve under your chairship, Ms McVey. I thank the hon. Member for Poole (Neil Duncan-Jordan) for securing a debate on a question that will shape every community in this country, including mine. I will try to focus on topics that have not been covered, including infrastructure. In my Mid Dunbartonshire constituency, AI has recently been brought to the forefront of local debate. At Westerhill, near Bishopbriggs, a developer has proposed a 300 MW AI data centre with 500 MW of battery storage—nationally significant infrastructure on our doorstep. My constituents’ response has rightly been one of caution—not hostility and not blind welcome. They have serious concerns and questions: “What jobs will this bring, and for whom? What happens to the environment, to the grid, to water and to energy bills?” Developments on this scale must be good neighbours and deliver community benefit that people can see, as well as skills, local employment and investment in the places that host them—not just high fences and, as the hon. Member for Dewsbury and Batley (Iqbal Mohamed) said, profits that are disseminated at a distance. That is how they will be judged by the communities who host them. This is an international issue, and it has so far been dominated by US policy. Last month, the United States showed the first sign that it may not see us as being part of its team when it comes to AI access. On 12 June, the US Government ordered one of the US’s leading AI companies to cut off access to its two most advanced models for anyone who is not a US citizen—the first time that export controls have ever been applied to AI models. Access for British businesses and researchers was switched off overnight not because of anything we did, but because Trump decided to cut us off. That is not a partnership; a vital technology that can be withdrawn at another Government’s whim is a chain around Britain’s neck. We cannot build our public services, our security and our economy on systems that can be cut off at a whim. It is clear that the UK must build its sovereign capacity—the data centres, skills and software needed to ensure that we are not locked out of the AI race. Communities such as mine, which are being asked to host the infrastructure of that future, should share fairly in the rewards, and we must build public trust. The Liberal Democrat approach would make that possible, based on a cross-sector, principle-based framework to give innovators certainty and the public confidence, and to give transparency wherever AI is used in the public sector, so that people can understand the impact it is having on their lives.

It is a pleasure to serve under your chairmanship, Ms McVey. I wholeheartedly thank the hon. Member for Poole (Neil Duncan-Jordan) for securing the debate. Its richness shows that we need also to have it in the main Chamber, because the Government need to tackle this issue head-on. Before I dive in, I refer Members to my entry in the Register of Members’ Financial Interests. The Government have done some work, but it still seems that AI policy is done by announcement, rather than by taking a real, strategic overview of how it will power our society, and there is still no sign of an AI Bill. As has been talked about today, AI is everywhere, and the Government must tackle it head-on. They need to see that it is a tool that can be used to benefit our society or that can lead to our society’s demise.

On the demise of society, AI can fuel the threat of disinformation, which can inflame tension and distort debate. We have seen that in Epsom and Ewell. Does my hon. Friend agree that AI must be regulated and that the Government must bring forward the AI Bill?

Absolutely, and that has been at the core of a lot of our debate today. AI can drive inequalities, break down our society, security and safety, and take away jobs, or it can be used to tackle our biggest problems, empower individuals and drive growth. AI is a bit like the electricity of today—a supercharged version of electricity—and we need to think about where we are going to direct it. At some point, we started using electricity to light our cities and to power our homes, businesses and industries; now we need to think about where we will direct AI, and how we prepare people for that future. The hon. Member for Poole and others talked about the future of our economy, tax and wealth, and there are key examples that highlight how important that is. We have started talking about donkeycorns; Members may have heard of unicorns, but donkeycorns are billion-dollar companies with one or two people. There is a legend that OpenClaw was a one-person company that was sold for £1 billion. That is a real shift in our economy, and we need to tackle it head-on. As Liberal Democrats, we believe in embracing that progress and change, but we also believe in empowering individuals. When it comes to AI, we thought about tackling that big picture, which is why we launched a project called “Everyday AI”. We brought together over 100 global stakeholders from industry and civil society—to look at a Liberal vision for AI and how Britain can thrive in the age of AI. We looked at backing British innovation, empowering people and building trust, and building infrastructure for success. As the right hon. Member for East Hampshire (Damian Hinds) said, no one knows where we are going. As a Member of Parliament and a politician, I often feel it is difficult to sound neither utopian nor dystopian about this issue. However, it is crucial that we ask the right questions, and keep on asking them, to make sure we are going in the right direction.

Will the hon. Lady give way?

I am going to advance, because I want to hear from the Minister. Crucially, we need to lean into the opportunities and challenges, and I will share some of our proposals, which I would be happy to discuss with the Minister afterwards. When it comes to leaning into the opportunities, it is about helping start-ups and small and medium-sized enterprises to thrive with AI, with AI procurement guides, skills tax credits and capital expenditure reliefs for AI software and adoption. We need to look at supporting our university spin-outs and the incredible research in this country; quintupling proof of concept funding; having a standardised national framework for negotiating IP between universities and staff; and moving towards full IP ownership by researchers and research teams. We have to help to grow these things here in the UK. As mentioned by the right hon. Member for Oxford East (Anneliese Dodds) and the hon. Member for Milton Keynes Central (Emily Darlington), we must use Government procurement to support UK tech. It has to be a catalyst for British sovereign tech. We are calling for a trial increase to 30% for the social value in procurement contracts. That would align with our strategic goals on digital sovereignty and AI, while maintaining open borders. When it comes to tackling the challenges head-on, we need to look seriously at the cyber-security threats that have been talked about today. The hon. Member for Poole talked about the impact of big tech companies, and Mythos was a wake-up call. We hear that China is now developing such models, but I am most worried about those we are not hearing about. We need to prepare for the future of work and fuel our creative industries, because we are nowhere without them. When it comes to NEETs, we have to realise that we are at a squeeze point, because companies are looking at how AI is bringing down costs. However, I would gently add that this is not the time to increase the cost of employment; that will not tackle the employment question. We need to look at establishing safety by design in AI systems and models. We must tackle harm from chatbots through a harm-based framework, but enforcement is crucial. The AI Safety Institute is now the AI Security Institute. It is such a shame that that focus has been diminished, because we should be looking at AI ethics, as well as safety and security. We have to ensure that AI systems are not used for harm. Members talked about AI fraud, which is a massive, serious problem in the UK. Revolut told us that the UK accounts for 22% of its customer base but 68% of its fraud costs. We are not tackling that massive problem properly, and AI is supercharging it. We are therefore calling for an online crime agency and AI labelling. We are also saying that social media companies, where a lot of AI-enabled fraud starts, should be financially liable for it. We must tackle the energy question for AI datacentres and their role in the community, which my hon. Friend the Member for Mid Dunbartonshire (Susan Murray) mentioned. We have to make sure they add value to our communities, and the AI Energy Council must tackle those issues head-on. On public services, we need to make sure a person stays in the loop. Ultimately, we need to prepare for a future that is arriving very quickly. Underpinning that, we are calling for a national people strategy. Things are changing very quickly, so we need a single, coherent framework, from primary school to the workplace, with regional input. We must realise that the skills we need for an AI future are not just technical skills, but critical thinking and analysis. We must also look at a digital sovereignty strategy, as many Members have said. We need to look at our tech stack to ensure that we do not have a single point of failure. A resilient tech stack would be good not just for our security and for businesses, but for our tech and our economy, so we absolutely need to look at that. We cannot give up on frontier models; we need to start looking at them now. A point was made about proportionality in models. We do not all need massive, complex models that require massive amounts of energy to solve most of our problems. Most companies probably need simple models. They can be UK models, but if we do not start now by backing and investing in them, we will be left behind very quickly. Finally, we need to move quicky on smart regulation. The Financial Conduct Authority has a fantastic approach to fintech that looks at output and standards. We need a flexible approach to AI, because things are moving quickly. We also need a digital Bill of rights. We need to unlock the benefits by addressing the challenges head-on. According to the Ada Lovelace Institute, 91% of the public feel that it is important that AI systems are developed and used in ways that treat people fairly, and almost 90% are calling for an independent regulator for AI. Will the Minister listen to those concerns, take back the many great ideas that have been raised today and speak to his colleagues across Departments? AI is an extremely powerful tool, and we must embrace it. The Government must tackle this head-on. Where will they let the power be focused? If anyone is unsure how quickly AI is going to change, I am repeatedly reminded that this year is the slowest we will ever see in AI development. If a plane going around the world is 1° off, it will be 500 miles off course. We need to decide now the direction we are going in with AI. We need to look at that holistically. We are here to support that Government work and make sure it empowers society, so that Britain can really be the leading nation on trustworthy, innovative technology.

It is a pleasure, as always, to serve under your chairmanship, Ms McVey. I thank the hon. Member for Poole (Neil Duncan-Jordan) for bringing forward this interesting debate, and all the Members on both sides of the House who have taken part. I recognise many of those who have participated in many debates such as this and who will, I am sure, participate in many more to come. I will start by thanking my right hon. Friend the Member for South Holland and The Deepings (Sir John Hayes)—I wish I had attended his lecture in 2000, not least to have had the experience of being in a room where everyone disagreed with him, which must have been rather unique—for making an important point about how slowly Parliament regulates and responds to change. My right hon. Friend the Member for East Hampshire (Damian Hinds) made a similar point about our responsiveness to this rapidly evolving sector. I also very much welcome his point about education and the challenges involved, particularly given his experience in the sector. However, more than anything, I welcome his point that nobody knows what the future outcomes will be—a point I shall come back to a bit later. I want to focus the bulk of my remarks on employment. I share the desire of the hon. Member for Poole to look at how to support and enhance employment through the AI transition, and we do need to have more such discussions. However, I am saddened that one core proposal from Labour Back Benchers is a plan on who or what we should tax now. I regret that the hon. Member’s plans for an employment levy, which he is very passionate about and has written about previously, would not work to protect jobs—protectionism never does. It would simply put a brake on the ability of organisations, large and small, to adapt to AI, and leave us behind our competitors, losing jobs across all sectors. The history of protectionism is, sadly, frequently a very tragic one. With each technological intervention and change, there are often calls for protectionist approaches, and when they come in—

Will the hon. Member explain how this will work, then? Human beings at work pay national insurance, income tax and pension contributions, and then they can support their families. AI agents and AI-powered robots do not pay any of that—there are no tax receipts for the Government—so how will Government fund public services in that scenario?

I thank the hon. Member for his challenge regarding a future hypothetical that has yet to be realised. I think I will be able to address some of his points, but I think the premise he is setting is a bit beyond the scope of this discussion. However, I can give some concrete proposals for protecting human jobs. As a starter for 10, rather than inventing a new tax, we could start with some old ones, and rather than inventing new regs, we could start with some newish ones. The Labour Government’s approach has been to make it as difficult as possible to employ human labour by increasing national insurance contributions from employers; making it more difficult for younger people to compete in the workplace by increasing the minimum wage; and increasing employment regulations, making it a greater gamble for businesses to take on new starters. To support jobs in the UK, we should, before looking at AI, look at the Labour Government’s policies, which are uniquely designed to make it more and more difficult to employ a person over a machine. We are right to think about the workforce implications of the adoption of AI, but the future is always unclear and uncertain; that is the challenge of working in this sector, and I am very mindful of that position of humility as I move on to the following, more specific remarks. Other technological disruptions, while displacing jobs, have ultimately led to more jobs through elasticity of demand. As costs decrease through automation, demand increases. That is the story of what happened with mass manufacturing in the automotive sector: to begin with, people were concerned that mechanised manufacturing would reduce jobs in the sector. Of course, the car sector exploded because the cost of a car went through the floor, and as a net result there were more jobs in the automotive sector. The issue was that they were different jobs; it is more of a transition than a replacement.

I call the Minister, mindful that I would like to give some time to the hon. Member for Poole to wind up the debate.

It is a pleasure to serve under your chairship, Ms McVey. I congratulate my hon. Friend the Member for Poole (Neil Duncan-Jordan) on securing the debate, which speaks directly to the kind of country and society that we want to create for everyone we represent, and our friends and family—the type of country and society we want to live and thrive in. The Government recognise that AI is, as Members have said, already changing our economy, the delivery of public services, our expectations of each other and the ways in which we interact and engage in debate. For me, it is not merely the latest thing, or the something new and shiny that is so attractive to politics and politicians; in my mind, it is equivalent to a new atmosphere in which we now all expect to feature and play our part. We see the potential of AI in everything from scientific discovery to better public services and stronger economic growth. The Government have a clear responsibility to make sure that the change at the heart of this debate reinforces rather than weakens the fabric of our society. That includes standing up to big tech firms when we deem it necessary—examples will follow—when they facilitate harm to UK citizens. As so many Members have powerfully made the argument for, it also includes developing sovereign capability—determining what comes next and honing it for best benefit. As we bring this debate to a close, it is important that we continue to consider those questions. We must collectively decide how we build a better society in an AI-enabled world. AI must serve us, and not the other way round. The Government recognise the pressure to internationally compete on AI. That means we need to actively shape the transition towards it to secure benefits for the UK while managing risk or, as my right hon. Friend the Member for Oxford East (Anneliese Dodds) put it, exert the domestic control that we need. That is why we are working hard to ensure that the UK has the capability, access and influence it needs in the technologies that will shape our economy, public services and national security. It is why we established the sovereign AI unit with £500 million to enable the Government to back promising UK firms. The UK already attracts the third highest private investment in AI.

The Teesside region has the potential to bring nationally significant investment in new AI data centre infrastructure. We welcome that investment and we are open for business, but will the Minister commit to meet colleagues at the Ministry of Housing, Communities and Local Government to ensure that the governance arrangements around the Teesworks site attract business, rather than deter it?

If I am considered worthy to meet in a few weeks’ time, I will be very happy to do so. This is not about self-sufficiency or about turning away from trusted international partners; it is about ensuring that where the UK has genuine strengths—in research, talent, creativity and innovation—we can translate them into long-term economic benefit. Significant private investment is now flowing into areas that would have seemed like science fiction even a decade ago, including therapies that slow or reverse biological ageing and AI-accelerated drug discovery that could compress timelines for curing major diseases. If realised, such developments would profoundly reshape expectations of health, work and the role of the state. That is why the Government have already committed to ambitious plans to transform services and digitise government, as set out in the road map for modern digital government published earlier this year.

I hear what the Minister says, but there are no specific laws governing the standards of AI, there is no dedicated regulator to deal with AI development, and there are no sanctions on companies that knowingly use AI for malevolent purposes. Those things are all needed now.

The Government’s principle is to regulate at the time of deployment or at use. I understand the right hon. Gentleman’s argument, but it is important that we strike the right balance between innovation and regulation. Of course, as I will come on to discuss in more detail, where there have been examples of harm we must call that out and call in those who caused them. The Government have already committed to ambitious plans to transform services. For a radiologist, that means access to the best tools for detecting cancer—tools that can save lives and cut waiting lists in our NHS. The public sector is where it is perhaps most crucial that the Government ensure that we embed frameworks, standards and guidance to ensure that AI is deployed in a way that is fair, transparent and accountable. We are doing so with the data and AI ethics framework, the AI playbook, and transparency standards that ensure the public can understand how systems are used. We are also supporting UK firms that provide third-party AI-assurance ecosystems, demonstrating that AI systems are safe and reliable as well as creating new jobs. That is critical not just for risk management but for building businesses, building confidence in those deploying AI, and building trust in the systems we use. To the points that were made about the job market, AI has the potential to enhance job roles and the performance of each of us at work, but it is necessary and reasonable, as my hon. Friend the Member for Poole pointed out, to distinguish the jobs that are most at risk—those least likely to be enhanced and most likely to be replaced. That is a very reasonable proposition. We recognise that one of the primary areas where people will interact with AI is in the labour market. To zone in on labour markets, and the issue of young people and entry-level jobs, an estimated 12.8% of all people aged 16 to 24 in the UK are not in education, employment or training. That goes well beyond data-entry or access-level jobs, or a potential shift of culture in relation to entry-level jobs. To date, there is little conclusive evidence that AI is reducing employment. However, the International Monetary Fund estimates that around 70% of UK workers are in exposed occupations. Around half of those exposed workers are in high complementary roles, where AI will enhance rather than replace their work. The Government are committed to supporting the worker experience through the AI transition, and have a proud record of advancing workers’ rights. To that end, earlier this year the Government established the AI and the future of work unit, bringing together expertise from across Departments and industry to monitor how AI is affecting jobs, wages and opportunity in real time. We are putting in place the institutions, data and cross-Government co-ordination needed to ensure that AI transforms the labour market, that workers are supported and that opportunities are widely shared across society. We are also investing in people, because the opportunities of AI will be realised only if the workforce are equipped to harness them. That is why we have launched one of the most ambitious skills programmes anywhere in the world. Our AI skills boost programme is already delivering results. In an answer given to the House last week, I referred to the early careers funding and work on curriculum reforms to come, such as an AI-inclusive GCSE in computing. This is about ensuring that workers at every stage of their career, and whether in small businesses, large firms or public services, have the tools that they need to thrive in an AI-enabled economy. AI will have—is having—seismic impacts on our working lives and productivity, and we must embrace the democracy this gives us. We must democratise access to AI to ensure that communities across the country reap the benefits. Central to that effort is ensuring that support is delivered to the graduates and young people entering the world of work for the first time. Much has been made of the rising level of NEETs. Programmes such as TechLocal are creating new jobs, traineeships and professional training, while scholarships and fellowships are supporting the next generation of researchers and innovators. Nobody is waiting 25 years to tackle the threats, however. AI harms are very real, and we do not sit idly by when it comes to tackling them. We are not standing still. The AI Security Institute is conducting world-leading research to understand the capabilities and impacts of advanced AI. In response to the generation of vile and degrading non-consensual sexual deepfakes on Grok earlier this year, the Prime Minister himself, and the Government, stood up to Grok and X and won that fight. We have criminalised the creation of non-consensual sexual deepfakes and legislated to require platforms to remove them. Separately, we have banned nudification apps and criminalised AI tools designed to create sexual abuse material, as well as learning the importance of sovereignty when considering the recent switching off of the latest Anthropic AI tool. I was asked a specific question about the social media ban. I was a proud early adopter of the policy and I am pleased to say that it remains the Government’s intention to update the House before recess. The UK is proactively responding to the AI revolution—we are shaping it. AI has extraordinary potential to improve healthcare, accelerate scientific discovery, personalise education, strengthen public services and drive economic growth. We will harness the opportunities across every region and community to shape UK society for the better for all.

I thank all Members for their contributions to an excellent debate that showed we need to have more of this discussion. I will make some very quick points. First, we need to act with speed, as I think was accepted by a number of Members. As the right hon. Member for South Holland and The Deepings (Sir John Hayes) mentioned, we cannot wait 25 years. We also need clear democratic oversight of what is happening. We cannot allow people we do not know and do not control to make our futures for us. We have to be in control of our own destinies. That is very important and will require global regulation. We will have to work with other authorities across the globe to come to the necessary arrangements. Finally, the protections we will need must ensure that the next industrial revolution, which is already here in effect, benefits everyone and not just a few. Motion lapsed (Standing Order No. 10(6)).