AI safety evaluators gain prominence as labs face internal and external pressure

As major AI labs like Anthropic and OpenAI move to embed third-party safety evaluators, the industry faces unresolved questions regarding funding, independence, and the scope of access for these auditors.