← Back to Model Beat
Research·Sep 6·all news from September 6, 2026

OpenAI Top Scientist Urges ‘Extreme Caution’ With Pace of AI

Leopold Aschenbrenner, a researcher formerly associated with OpenAI and Anthropic, has left Anthropic following disagreements over the company's internal safety protocols. He stated that current security measures at leading AI labs are insufficient to prevent the potential theft of sophisticated models by foreign intelligence agencies. This departure highlights an ongoing debate within the industry regarding whether development speeds currently outpace the implementation of robust safeguards against national security threats.

Covered by 38 sources · 60 articles

Related stories

ResearchMeasuring AI capabilities in intelligence targeting and conventional weaponsSep 9 · 60 sourcesResearchResearch acceleration: The view inside OpenAISep 6 · 5 sourcesResearchAI-Discovered Drug Reverses Aging Markers in Study, Biotech SaysSep 7 · 4 sourcesResearchPaul Christiano joins OpenAI Foundation BoardSep 9 · 3 sources