UN touts panel for 'human control' of AI at global summit

NEW DELHI, Feb 20 (Agencies): A UN panel on artificial intelligence (AI) will work towards "science-led governance", the global body's chief said on Friday as leaders at a New Delhi summit weighed their message on the future of the booming technology.

However, the US delegation warned against centralised control of generative AI, highlighting the difficulties of reaching consensus over how it should be handled.

The flip side of the gold rush around AI is a host of issues from job disruption to misinformation, surveillance and the heavy electricity consumption of data centres.

"We are barrelling into the unknown," UN chief Antonio Guterres told the AI Impact Summit in New Delhi. "The message is simple: less hype, less fear. More facts and evidence."

To cap the five-day summit, dozens of world leaders and ministers are expected to deliver on Friday a shared view on the benefits of AI, such as instant translation and drug discovery, but also the risks.

It is the fourth annual global meeting focused on AI policy, with the next to take place in Geneva in the first half of 2027.

Tens of thousands of people from across the sector visited the vast summit and expo site, with some complaining of poor organisation including crowded and chaotic entry and exit points.

Police detained a group claiming to be members of the youth wing of the opposition Congress party who staged a shirtless protest at the venue against Prime Minister Narendra Modi on Friday.

- 'Science-led governance' -

Guterres said the UN General Assembly had confirmed 40 members for a group called the Independent International Scientific Panel on Artificial Intelligence.

It was created in August, aiming to be to AI what the UN's Intergovernmental Panel on Climate Change (IPCC) is to global environmental policy.

"Science-led governance is not a brake on progress," and instead allows a "move from rough measures to smarter, risk-based guardrails", Guterres said.

"Our goal is to make human control a technical reality -- not a slogan."

White House technology adviser Michael Kratsios, head of the US delegation, warned that "AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralised control". "As the Trump administration has now said many times: We totally reject global governance of AI," he said.

The United States did not sign last year's summit statement and it released its own joint declaration with India on Friday afternoon.

It said the two countries would "pursue a global approach to AI that is unapologetically friendly to entrepreneurship and innovation".

"Both sides pledge to adopt... mainstream regulatory regimes that advance technological innovation and promote investment," the bilateral statement said.

The Delhi gathering is the largest AI summit yet, and the first in a developing country, with India taking the opportunity to push its ambitions to catch up with the United States and China.

India expects more than $200 billion in investments over the next two years, and US tech titans unveiled a raft of new deals and infrastructure projects in the country this week.

The broad focus of the summit, and vague promises made at its previous editions in France, South Korea and Britain, could make concrete commitments unlikely.

Sam Altman, head of ChatGPT maker OpenAI, has called for oversight in the past but said last year that taking too tight an approach could hold the United States back in the AI race.

Meanwhile, dozens of world leaders and ministers were expected to deliver on Friday a shared view of how to handle artificial intelligence, wrapping up a five-day summit focused on the technology.

It comes a day after OpenAI chief Sam Altman told the meeting in New Delhi that the fast-evolving sector needs regulation "urgently".

Frenzieddemand for generative AI has turbocharged profits for companies, while also fuelling fears about the impact on society and the planet.

AI summit statement delayed

to 'maximise' signatories

Dozens of national delegations at an artificial intelligence summit in India will issue their statement on how the world should handle the technology on Saturday, a day later than expected, the host country said.

"There is huge consensus on the declaration. We are just trying to maximise the number," India's IT minister Ashwini Vaishnaw told reporters at the AI Impact Summit in New Delhi on Friday.

"The declaration and its contours will be shared transparently tomorrow," he said, adding it had more than 70 signatories so far but he hoped the figure would cross 80.

Vaishnaw declined to give details of what the statement would say as he thanked participants of this week's event that was attended by tens of thousands of people, including world leaders and tech CEOs.