An independent assessment by safety evaluation group METR found Anthropic is developing AI capabilities at twice its historical rate. Evaluators attribute roughly 80 percent of the acceleration to internal AI use, though Anthropic heavily redacted supporting data.
FireBrief
FireBrief