AI Models & Agents
METR: AI may be speeding Anthropic’s model progress
An assessment using internal data estimates roughly twice the past rate of progress on well-specified tasks, with substantial uncertainty.
Researchers review abstract charts illustrating AI model performance and human-AI collaboration.
Photo: Kronos News
An independent assessment by METR estimates that, as of October 2026, Anthropic was improving its models’ capabilities on well-specified tasks at about twice the rate seen from March 2024 to October 2025. METR says its review, conducted from August to October, drew on internal company data and staff interviews. [1]
One measure METR examined found Anthropic model benchmark scores growing at 1.9 times their earlier rate since October 2025, with a 95% confidence interval of 1.5 to 2.8 times. METR believes AI use likely accounts for a majority of the current acceleration, but says that attribution is uncertain; it also warns that benchmark results may not reflect broader real-world capabilities. Anthropic redacted much of the evidence METR considered most informative, limiting what readers can independently assess. [1]
Editorial notes
Transparency note
AI assisted drafting. Human edited and reviewed.
- AI assisted
- Yes
- Human review
- Yes
- Last updated
Risk assessment
METR reports substantial uncertainty in its estimates, and Anthropic redacted much of the evidence most relevant to its conclusions.
Sources
Related stories
View all
Topics
About the author
Kronos News Desk covers ai models & agents and editorial analysis for Kronos News.