Anthropic grants METR access to AI models for safety and transparency push

Anthropic has granted METR access to its AI models to enhance safety and transparency. This strategic move allows for a comprehensive AI safety review of the technology. Current projections suggest a 46.5% probability that Anthropic will possess the best AI model by November 2026. By providing this access, the company aims to push for greater transparency within the industry, ensuring that its advanced models are scrutinized by external safety experts to mitigate potential risks.