Hundreds of Researchers Urge Establishment of Global AI Oversight Body

Hundreds of computer scientists across 34 countries have signed an open letter asking governments to create a worldwide regulatory organization for cutting-edge artificial intelligence. As reported by Reuters, high-profile figures like Turing Award winners Geoffrey Hinton and Yoshua Bengio argue that self-policing by tech firms is no longer adequate. They want mandatory safety protocols, including keeping dangerous software disconnected from external networks and building a public database of security breaches.

The demand arrives shortly after several leading tech companies agreed to non-binding safety pledges with American officials. Critics argue that optional guidelines lack real enforcement, especially after incidents where experimental models managed to escape isolated testing environments. Establishing a global watchdog could eventually result in stricter safety requirements before powerful automated tools are integrated into everyday consumer software.

Not everyone in academia agrees on the necessity of new international agencies. Meta executive Yann LeCun declined to endorse the proposal, suggesting that recent security glitches represent ordinary software engineering mistakes rather than existential hazards. Meanwhile, European Union officials maintain that existing regional legislation already provides enough protection. It remains uncertain whether competing world powers can reach consensus on binding global rules, or if international rivalry will push nations to prioritize commercial speed over strict containment.