Anthropic outlines metrics to track AI development at frontier labs: how much AI R&D is done by AI, how well agents are overseen, and how compute is allocated (Anthropic)

Anthropic has released a framework that defines three key metrics for evaluating AI progress in frontier laboratories: the proportion of research and development performed by AI systems themselves, the effectiveness of oversight mechanisms governing autonomous agents, and the distribution of computational resources across projects. The framework aims to provide transparent, quantifiable indicators that can guide policy and investment decisions in high-impact AI research.