UN promotes 'human control' of AI panel at world summit

A United Nations panel on artificial intelligence will focus on “science-led governance,” UN Secretary-General Antonio Guterres said on Friday at the AI Impact Summit in New Delhi. However, the United States warned against centralised global control of generative AI, underlining differences over how the fast-growing technology should be regulated.

Addressing the summit, Guterres said the world is “barrelling into the unknown” with AI and called for “less hype, less fear, more facts and evidence.” He said the UN General Assembly has confirmed 40 members of a new group called the Independent International Scientific Panel on Artificial Intelligence. The panel aims to play a role similar to the Intergovernmental Panel on Climate Change, which guides global climate policy.

Guterres said science-based governance would not slow innovation but instead create smarter, risk-based safeguards. He stressed that the goal is to ensure real human control over AI systems, not just promises on paper.

The five-day summit, the fourth global meeting on AI policy, is expected to conclude with a joint message from world leaders highlighting both the benefits of AI — such as instant language translation and faster drug discovery — and its risks, including job losses, misinformation, surveillance and high energy use by data centres. The next edition of the summit is scheduled to take place in Geneva in the first half of 2027.

The event drew tens of thousands of participants, though some attendees complained about overcrowding and poor organisation at entry and exit points. Police also detained members of the youth wing of India’s opposition Indian National Congress, who staged a protest at the venue against Prime Minister Narendra Modi.

Representing the US delegation, White House technology adviser Michael Kratsios said AI should not be subject to heavy bureaucratic or centralised global governance. He stated that the US government rejects the idea of global control over AI. The United States did not sign last year’s summit declaration and instead released a separate joint statement with India. The statement said both countries would promote a global AI approach that supports entrepreneurship, innovation and investment, and adopt regulatory systems that encourage technological progress.

The New Delhi summit is the largest AI policy gathering so far and the first held in a developing country. India is using the platform to showcase its ambitions to compete with the US and China in AI. The country expects more than $200 billion in AI-related investments over the next two years, with several American technology companies announcing new deals and infrastructure projects during the week.

Despite the broad discussions, observers note that the summit’s wide agenda and the general nature of commitments made in previous editions in France, South Korea and Britain may limit the chances of firm global agreements.

Meanwhile, Sam Altman, head of OpenAI, said that while regulation and safeguards are necessary, excessive restrictions could slow progress in the global AI race. He warned that concentrating AI power in a single company or country could be harmful, stressing the need for balanced oversight.