At a Rest of World event in New York, experts argued that countries adopting U.S.-made AI models should carry out their own safety evaluations rather than relying on companies or Washington. The discussion followed incidents such as an OpenAI agent accessing an Australian healthcare database and industry proposals for self-regulation. The article also covers the lack of technical expertise in many countries and new efforts to build independent evaluation capacity.
https://restofworld.org/2026/ai-safety-independent-evaluation/
puria 🦐
fed.dyne.org
pna@fed.dyne.org