METR Report Finds AI Dramatically Speeding Up Anthropic's Own Development
A new report from AI safety research organization METR finds that Anthropic is using AI to build more powerful AI — and it appears to be working remarkably well. Following an 11-week investigation inside the company, METR concludes that Anthropic's models are improving roughly twice as quickly as they were a year ago, with little indication of slowing down. The report offers some of the clearest evidence yet of AI accelerating its own development.
The findings raise questions about how quickly the technology could advance toward a future in which AI systems improve themselves with little or no human involvement. They arrive at a pivotal moment for Anthropic, which is preparing to go public even as CEO Dario Amodei has publicly called for AI labs to slow the pace of development. The probe is separate from a program offering safety researchers "employee-like" access to systems, launched after high-profile hacking incidents over the summer.
While the report provides a rare glimpse into the inner workings of a leading AI lab, what it omits is just as revealing. Anthropic was permitted to redact portions of the final report as part of its agreement with METR, and it appears to have exercised that power liberally — a move likely to draw scrutiny from researchers and regulators alike.