METR audit: Anthropic model R&D accelerated by about double
The story so far
After an internal investigation of 11 weeks at Anthropic, the AI Safety Research Institute METR released a report stating that the capability improvement rate of Anthropic's models is about twice as fast as a year ago, achieving about two years of historical progress in the past year. METR claims that the acceleration mainly comes from AI-assisted human researchers, not from AI working independently, but warns that 'there are already some signs that fully autonomous contributions will soon play a larger role.' In specific metrics, METR states that since October 2025, the improvement rate of model test scores has been about 1.9 times faster than before, and the annual growth rate of AECI has increased from about 15 points to 29 points; the true acceleration factor may be between 1.3 and 3. If OpenAI, Google, and xAI's models are included, the industry figure drops to 1.4, and METR believes this may be due to Anthropic falling behind competitors in 2025 and the chasing effect distorting the data. METR also believes that Anthropic may have crossed its own AI acceleration warning threshold on July 15 and accuses the public report of underestimating the acceleration rate and uncertainty. The verifiability of the report is limited: Anthropic has unilateral editing rights to non-public information and uses a large amount of it, METR claims it submitted a complete report of about 100 pages, but the 'most informative evidence we found' was edited out in the public version; Anthropic also did not grant unrestricted access to researchers, some data requests were rejected, and part of the computing power audit was canceled midway.
AI-written from the reports · Updated 6 days ago
Developments
3 developments
- 7 Nov, 20:30 · 1 reportsAI Acceleration ReportThe Verge · AI:Why a new report on Anthropic automating AI development is alarming researchers
- 7 Nov, 04:27 · 1 reportsAI dramatically accelerating AI researchHacker News · AI hot posts:AI is dramatically accelerating AI research
- 5 Nov, 16:00 · 1 reportsIndependent evaluation of Anthropic's AI R&D acceleration in October 2026METR:Independent Assessment of AI R&D Acceleration at Anthropic in October 2026
Coverage timeline
Follow the coverage to see every side of the story.
- The Verge · AIFeaturedWhy a new report on Anthropic automating AI development is alarming researchers
AI safety research organization METR released a report after an internal investigation of 11 weeks with Anthropic, stating that the capability improvement rate of Anthropic's models is about twice as fast as a year ago, achieving about two years of historical progress in the past year, and researchers are confident that the acceleration mainly comes from AI rather than computational power expansion.
- Hacker News · AI hot postsAI is dramatically accelerating AI research
Anthropic's internal report shows that its AI capabilities advanced approximately twice as much between October 2025 and October 2026 as in the previous year, mainly attributed to the acceleration of AI automation in R&D. The report indicates that to maintain the current doubling rate of AI capability improvement, researchers would need about a 10-fold increase in labor efficiency, which aligns with the internal model's 12-fold efficiency improvement reported by Anthropic employees. This accelerated trend may shorten the timeline for AI development to an earlier stage than most external researchers predicted.
- The DecoderFeaturedMETR finds Anthropic may have crossed its own AI acceleration warning threshold back in July
METR's internal audit of Anthropic shows that since October 2025, the rate of improvement in its model test scores has been about 1.9 times faster than before (from March 2024 to October 2025). AECI's annual growth rate has increased from approximately 15 points to 29 points, equivalent to completing two years of progress in one year.
- METRIndependent Assessment of AI R&D Acceleration at Anthropic in October 2026
METR evaluation shows that Anthropic's model capability improvement rate has roughly doubled since March 2024 to October 2025. Most of the acceleration is attributed to AI-assisted R&D, but there is significant uncertainty; the AECI benchmark score growth rate is approximately 1.9 times per year (95% confidence interval 1.5-2.8 times), suggesting about two years of progress in one year. Most of the key evidence has been redacted by Anthropic, making it difficult for the public to judge independently.