On 5 November 2026, METR published an independent high-access assessment of AI research-and-development acceleration at Anthropic. From 12 August to 29 October 2026, METR staff reviewed internal metrics, benchmark scores, surveys and staff interviews to judge how far AI is speeding up further AI development.
Most METR staff on the project now believe that, as of October 2026, Anthropic is likely improving the capabilities of its models on well-specified tasks at around twice the rate seen between March 2024 and October 2025. METR stresses high uncertainty and would not be surprised if the true figure were closer to 1.3× or 3×. Benchmark scores measured by AECI (Anthropic’s version of the Epoch Capabilities Index) have grown at an estimated 1.9× the pre-October 2025 rate (95% confidence interval 1.5–2.8×). METR is confident that the acceleration was not due to faster scaling of training compute. A majority of the current acceleration is judged likely attributable to Anthropic’s extensive use of AI to do R&D, though with substantial uncertainty; other factors likely played a large role in late 2025.
METR did not identify clear bottlenecks in algorithmic progress, training compute, experiment compute or data that would prevent further rapid progress over the next 12 months, and therefore sees a significant risk that further capabilities progress could lead to even faster acceleration—while noting that important bottlenecks may still have been missed. Anthropic redacted a majority of the evidence METR found most informative for its conclusions, describing the redactions as aimed at removing confidential information that could accelerate other AI developers. METR warns that this leaves the public more dependent on METR’s judgment than would be ideal.
Note from Evangelii Network Japan: This comment is our own and not part of METR’s assessment. Independent scrutiny of how AI accelerates AI development is a question of human dignity and the common good, explored further in our paper on Catholic AI and human dignity.
