TechAnthropic reports Claude now leads 26% of its internal AI R&D work
Anthropic published metrics showing Claude models lead 26% of the company's AI research and development tasks, with AI collaborating on or leading over 90% of R&D work overall. The company introduced a new 'R&D Automation Index' to track progress toward recursive self-improvement.
TLDR
This provides concrete public data on how close frontier labs are to models meaningfully accelerating their own successors, raising safety, control, and pace-of-progress questions. The rapid jump from ~1% in March signals accelerating AI-assisted model development and is seen as a transparent benchmark that other labs might follow, amid ongoing debates on AI development speed and recursive self-improvement.
Combined views
24.5K
2 Sources, first seen 17h ago
Anthropic reports Claude now leads 26% of its internal AI R&D work
Anthropic published metrics showing Claude models lead 26% of the company's AI research and development tasks, with AI collaborating on or leading over 90% of R&D work overall. The company introduced a new 'R&D Automation Index' to track progress toward recursive self-improvement.