Anthropic says Claude leads 26% of its measured AI research work

Anthropic estimates Claude reached the 'leads' level for 26% of its measured AI R&D work in August, up from under 1% in February, with over 90% at 'collaborates or above' and zero full autonomy. About 30,000 agents ran at once on its largest internal platform, with roughly one in 47,000 of over a billion decisions blocked. CEO Amodei has called for pacing AI progress and outside evaluators.

• Anthropic CEO says AI could cure most diseases within a decade2026-09-24
• AI anxiety grows; therapists advise watching daily impact2026-09-22
• NYT: concerns that AI could kill us reach public attention2026-09-17
• Trump says he uses AI ‘for a lot of things’ without details2026-09-14
• Dreamforce draws 43,000+ amid AI debate2026-09-14