I thank the hon. Member for those points. I have already said that we must embrace AI and what it does for us. We are not here to stop progress, but my point is that the Government need to build in regulatory rights and protections.
The benefits of this new technological revolution must be shared by everyone, not just an elite few. I do not claim to have the answers to a challenge of such enormous magnitude—I look forward to hearing hon. Members’ thoughts in a few moments’ time—but a starting point must surely be guaranteeing support to those sectors and communities that will be most affected by the threat and reality of economic displacement. That means strengthening our collective social security net and seriously considering the role that a universal basic income might play in ensuring a decent standard of living in a labour market increasingly characterised by job scarcity. It means investing in skills and lifelong learning, ensuring that workers whose employment is lost to AI have the opportunity to find well-paid and similarly rewarding work.
In any democracy we have to recognise that technology is never ideologically neutral. Every technological system reflects the interests and biases of its creators and funders. Our challenge is to ensure that AI technologies reflect a multiplicity of voices, including those of workers, and not just in their application but in their conception and design as well. I hope we will continue to discuss how we can achieve that.
A people-focused approach to AI must also mean doing more to guarantee the rights of those workers who are already working alongside artificial intelligence and related technologies in their workplace. The AI working group set up by the Trades Union Congress surveyed thousands of workers in producing its report on the
worker experience of AI and associated technologies. It shows vividly how workers are increasingly managed by machines, how their rights and autonomy are being steadily eroded, and how automated processes often perpetuate human prejudice when making decisions on employees’ performance, hiring and promotions.
The Government’s response was set out in the Department for Science, Innovation and Technology’s recently published AI White Paper, which advocates a light-touch approach and effectively leaves the market to regulate itself. Although Ministers have devised five fundamental principles that should inform the adoption and use of AI in workplaces, they do not intend to place those principles on a statutory footing. Instead, the implementation of those principles will be left to underfunded and overstretched regulators, such as the Information Commissioner’s Office and the Equality and Human Rights Commission.
That contrasts starkly with the models adopted by other developed economies. The European Union’s Artificial Intelligence Act is likely to be one of the most comprehensive pieces of legislation ever passed on this subject, while California—the very centre of global technology innovation—is preparing to implement measures to protect the privacy and civil liberties of workers. These measures include a new office for AI, with the authority to guide the development of new automated systems, as well as statutory restrictions on the use of automated decision making in the workplace.
The proposal set out by the TUC’s AI manifesto, copies of which I have brought to Westminster Hall for Members today, involves taking a very different position from that taken by the Government. Building on the existing framework of equalities legislation, it calls for a rights-based approach to manage the transition to AI that would strengthen equality protections, guarantee workers the right to human contact and require a human review of high-risk decisions that have been automated, and protect the right to disconnect for all workers. It is also absolutely right to acknowledge the need to listen to workers—their voices and their experiences—in managing this transition. It is essential that we recognise and value the role of trade unions as a vehicle for getting those voices heard.
It is for those reasons that the manifesto proposes a statutory duty for employers to consult trade union representatives before adopting AI and associated technologies. It is also why the manifesto urges employers to agree collective agreements with unions to govern the use of AI in the workplace.
Last December, when I questioned the then Business Secretary—the right hon. Member for Welwyn Hatfield (Grant Shapps)—on the merits of introducing a statutory duty to consult, he expressed interest and offered to meet me to discuss it further. I think the Minister present today will remember that, and I am interested to hear whether he and the new Business Secretary share the right hon. Gentleman’s interest.
Finally, the manifesto emphasises the fact that workers’ participation can be achieved only if workers understand the processes and technologies at work. In environments in which decisions are increasingly dictated by machines, people need to know, more than ever, what data is being held on them and how it is used.
I am aware that time is short and I look forward to hearing other hon. Members’ contributions. I will conclude my remarks by saying that on 17 May I will introduce a ten-minute Rule Bill that builds on the TUC’s important work and which I hope will bring us a bit closer to the rights-based approach I am advocating and which we urgently need. I ask any colleagues interested in supporting that Bill to speak to me after this debate.
2.44 pm