Sep. 15 at 8:14 PM
Elon Musk called on leading AI companies to conduct mutual safety testing of their models before releasing them publicly, speaking Monday at the All-In Summit in Los Angeles. Musk said xAI, OpenAI, Anthropic, Google, Meta and major Chinese AI companies should allow competitors to evaluate their models, arguing that outside rivals could identify risks that companies might overlook when reviewing their own systems.
The proposal comes after a weekend of heightened warnings about AI risks. Anthropic CEO Dario Amodei called for a slowdown in frontier-model development, a position that received support from Musk and OpenAI CEO Sam Altman. Concerns intensified after former Anthropic and OpenAI employee Jacob Coxon resigned and warned that leading AI labs were “playing with our lives.”
Musk acknowledged that rival AI labs have not accepted his peer-review proposal. Anthropic has separately suggested using third-party “embedded evaluators” to assess safety practices.
$GOOGL $META $SPCX