Anthropic has committed to using embedded third-party evaluators, while OpenAI would adopt the same arrangement. The proposed setup would give evaluators company badges, desks and laptops.
What Anthropic and OpenAI have committed to
Anthropic has committed to using embedded third-party evaluators, describing the arrangement as one strategy for pacing the AI frontier. OpenAI would adopt the same approach.
How the evaluator arrangement would work
Under the proposed arrangement, third-party evaluators would receive company badges, desks and laptops. OpenAI would follow Anthropic’s commitment and adopt the same embedded-evaluator setup.
The broader context around Anthropic
Anthropic’s stated mission is the responsible development and maintenance of advanced AI for the long-term benefit of humanity. The company also paid $1.5 billion to settle with authors.