UN chief Guterres and Barbados's Mottley to host a special AI safety summit in Bridgetown on Monday

The summit will open talks on Guterres's proposed multilateral AI risk-management framework with independent oversight; UN envoy Amandeep Singh Gill says it is "not a treaty, it's not regulation" but a basis for countries to build national AI safety capacity.
Why it matters
The meeting is the UN's attempt to give countries without their own testing capacity a common basis for AI safety work. Gill cites a call by Norway and Finland for control of frontier AI models, now signed by 32 states and the European Commission. About 150 participants are expected, including heads of AI safety institutes, a senior US diplomat and Chinese experts. Military uses of AI are excluded and no negotiated outcome is planned, so watch for pledges.
Key facts
- Bridgetown Special Summit on AI Safety, Barbados, Monday 23 November
- Co-hosted by António Guterres and Mia Mottley
- 135 leaders raised AI at this year's UNGA high-level week, up from 3 in 2017
- Norway–Finland call on frontier AI signed by 32 states and the European Commission
- About 150 participants; military AI not covered
- High-income countries have 500 incident response teams, low-income 6