METR finds Anthropic may have crossed its AI acceleration warning threshold in July
AIMETR, a nonprofit, audited Anthropic's internal benchmark data and found AI-driven progress roughly 1.9 times faster since October 2025 than before. METR says Anthropic may have reached its warning threshold on July 15, and that the company's public report understated both the acceleration and the uncertainty around it.
AIWhy it matters.
The audit compares Anthropic's public risk report with its unredacted version, showing how redactions and public statements can understate AI acceleration.
Loading source posts…
Source details
The Decoder · Maximilian Schreiner · the-decoder.comPublished · added here
Original title: METR finds Anthropic may have crossed its own AI acceleration warning threshold back in July