Last month, we launched a new series exploring how businesses can deploy AI in ways that serve people, strengthen society and support a more sustainable future. We began with sustainability and the principle of stewardship. This week, as the world’s attention turns to the UN’s AI for Good Global Summit, we explore another pillar of Blueprint’s Framework for Using AI: building trust and trusted relationships.

AI’s trust problem

From 7–10 July, policymakers, researchers and business leaders are gathering at the AI for Good Global Summit, the UN’s leading platform on AI. The premise is optimistic: that AI can be harnessed to serve humanity and solve shared global challenges.

However, while AI is advancing at remarkable speed, with new tools, capabilities and use cases emerging almost daily, public confidence is struggling to keep pace.

New research from King’s College London Policy Institute and Institute for Artificial Intelligence, highlights that just 24% of the UK public think AI is positive for humanity. Almost seven in ten (69%) are worried about the economic impact of AI-related job losses. A majority (57%) believe AI will lead to widespread unemployment, and more than one in five (22%) think it could eliminate jobs quickly enough to cause civil unrest.

These findings reveal a significant gap between the optimism surrounding AI and the confidence people have in the way it is being developed and deployed. If AI is to be a force for good, organisations must address not only what the technology can do, but how its benefits and risks are shared.

Reciprocity and responsibility

In our framework, trust is grounded in the principle of reciprocity: the idea that healthy relationships depend on honesty, fairness and mutual benefit.

This is particularly relevant to AI. Those building and deploying AI hold significant knowledge and influence; those affected by it often do not. Employers decide how work changes. Technology companies shape how systems are developed and deployed. Employees, customers and citizens live with the consequences.

Reciprocity suggests that the greater the power you hold, the greater your responsibility to use it for the benefit of others.

Who benefits?

Public concern about AI is often framed as fear of the technology itself. But the evidence suggests something more nuanced. People are less worried about what AI can do than about who will benefit from it.

Around six in ten people surveyed by King’s College London expect the gains from AI to flow mainly to wealthy investors and large companies rather than workers or wider society. When people believe the rewards will be captured by a few while the risks are borne by many, trust begins to erode.

This raises important questions for organisations adopting AI. Are the benefits being shared fairly? Are organisations being transparent about how AI is used and where its limitations lie? Are they designing with those most affected by disruption in mind?

The public’s message is not that AI should stop. It is that AI should be shaped and deployed responsibly. As AI continues to advance, organisations that earn trust will be those that use their knowledge and influence not only to create value, but to share it.