NewzAnthropic has granted the external evaluation organization METR employee-level access to its artificial intelligence models. This move is part of a broader initiative by the AI company to enhance safety protocols and external oversight of its systems.
By providing this level of access, Anthropic allows evaluators to conduct deeper testing and scrutiny of its technology. The partnership aims to strengthen the safety evaluation process for advanced AI models, reflecting an industry-wide push for increased transparency and third-party verification in AI development.
In-depth summary · AI, neutral