Recent weeks have seen a notable shift in public discussion about artificial intelligence.

Warnings from senior figures within leading AI companies, concerns raised by former researchers, growing calls for statutory regulation and increased scrutiny from policymakers are all pointing towards the same underlying challenge: AI capability appears to be advancing faster than society’s ability to govern it.

As a result, the debate is entering a new phase. Questions about innovation and adoption remain important, but they are increasingly being accompanied by deeper questions about accountability, oversight and the public interest.

Much of this discussion can be understood through a governance gap: a situation in which technological capabilities evolve more rapidly than the institutions, norms and frameworks designed to oversee them.

What makes recent developments particularly significant is that these concerns are no longer being raised solely by academics, civil society organisations or policymakers. Increasingly, they are also emerging from individuals and organisations at the forefront of AI development itself.

The debate is therefore becoming less about whether increasingly capable AI systems can be built and more about whether societies have the means to ensure those capabilities are directed towards outcomes that genuinely serve people.

Beyond risk management

Most responses to these concerns have focused, understandably, on regulation, oversight and risk mitigation. As AI systems become embedded in more areas of economic and social life, effective governance, transparency and accountability will become increasingly important.

But governance alone cannot answer a more fundamental question: what are we trying to achieve?

Regulation can help determine how decisions are made, but it cannot determine what those decisions should ultimately be directed towards.

This distinction matters because discussions about AI are often framed in terms of avoiding harm. While essential, this is only part of the picture. A focus on risk tells us what we want to prevent; it does not tell us what we are ultimately trying to create.

As AI becomes more powerful and pervasive, questions of purpose become unavoidable: what should AI be used for? How should its benefits be shared? What responsibilities do organisations have to those affected by AI-enabled decisions? And what does success look like in an AI-enabled economy?

These are not simply technical or regulatory questions. They are questions about values, priorities and the kind of future we want to build.

Our Framework for human-centred AI, challenges organisations to look beyond efficiency, productivity and competitive advantage and consider how their use of AI contributes to wider human flourishing – to consider not only the value created for their organisation, but also the value created for people, communities and society as a whole.

This perspective adds an important dimension to the current debate.

Much public discussion currently oscillates between two positions: optimism about AI’s growing capabilities and concern about its risks. But we believe there is a third overarching question that needs more debate: what should this technology ultimately be for?

Rather than focusing solely on whether AI should be accelerated or constrained, what kinds of social, economic and human outcomes do we as a society want AI to serve?

This is fundamentally a question of stewardship. How can increasingly powerful technologies be directed towards outcomes that expand human capability, support meaningful work, strengthen communities and contribute to long-term societal flourishing?

A debate about society, not just technology

The significance of recent developments lies not only in the specific warnings, regulatory proposals or political responses that have emerged. Their deeper significance is that they reflect a growing recognition that AI is not simply a technological issue.

It is also a societal one.

Debates about AI are increasingly becoming debates about the future of work, the distribution of power and opportunity, the responsibilities of business and the kind of economy we want to build to support a flourishing society.

For organisations, this creates both a challenge and an opportunity. Compliance with emerging regulation will matter. Effective risk management will be essential. But neither on its own is likely to be sufficient.

There is an opportunity for businesses in the UK to lead the way in this. Used well, AI can genuinely augment human capability, extend access to services, improve decision-making, free people for more meaningful work and help us to solve some of the most pressing social and environmental challenges we are facing as a society.

What would need to be true for UK businesses to realise AI’s full potential? The conversation must move beyond what AI can do and focus on how investors, regulators and policymakers can create the conditions for success. The priority must now be fostering an environment that enables us to build more trusted, resilient businesses that create long-term value for society, strengthen the UK economy, and deliver benefits for everyone.