The BaseResearchMETR
Review of the "Misalignment in high-stakes settings" section in Anthropic's Risk Report (August 2026)
METR publishes its external review of the misalignment section of Anthropic's August 2026 Risk Report. METR agrees the risk is low for the covered models but attributes this mainly to limited capabilities, and says the report's methods could not reliably detect higher risk. It recommends end-to-end red-teaming of monitoring.
Why it matters
The same thing, for different readers.
What it means for you
An outside reviewer of Anthropic's own risk report agrees the danger is low today, but says that is because the models are not yet capable enough, not proven safeguards.
For lawmakers
METR's review shows what an external check of a lab's safety case can and cannot do under limited access, and recommends that Anthropic publish four pieces of redacted information.
For researchers
METR finds the report's covert-capability benchmarks a poor proxy for evading Anthropic's actual monitoring, and calls off-distribution monitor reliability an unsolved research problem.
The question it raises
A lab's claim that its AI is safe is only as strong as the evidence outsiders may see, and METR says this review had limited access.
In their own words
Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.
From the source’s own text.
“Our best guess is that we agree with Anthropic that the risk of catastrophic harm from misalignment is low for the models covered in the report. However, this is primarily due to models’ limited capabilities, rather than safeguards and alignment training.”
“The covert capabilities benchmarks in the report are a poor proxy for models’ ability to evade Anthropic’s internal usage monitoring system.”
“Some of these incidents seem well-described as “unknown severe context-dependent misalignment,” a failure mode the report argues is “unlikely.””
“Conduct end-to-end red-teaming of monitoring systems, such as an exercise in which a well-resourced red team attempts to start a rogue internal deployment without being detected.”
How the finding was reached
In our words, from the source. Weigh the finding by how it was made.
- What was done
- An external review of one section of Anthropic's internal August 2026 Risk Report, written from the report alone and described by METR as relatively lightweight.
- What it found
- METR's best guess is that misalignment risk is low for the covered models, mainly because of limited capabilities. It doubts the report's method would reliably show higher risk.
- What it does not show
- METR says the review used only the report, with less access than Anthropic's later announcement describes. Anthropic could veto or redact; METR says no redactions were made and no conclusions changed.
- Status
- A report
The record
- Questions
- Can AI systems act beyond what people intended or allowed? · Are AI companies putting stronger safeguards into practice? · Who can demand evidence, delay a release or stop an unsafe deployment?
- People
- Nikola Jurkovic (Wrote it) · Tim Hua (Wrote it) · Beth Barnes (Wrote it)
- What we read
- We could read only part of the source.
- Found by
- The Radar, run of October 21, 2026, via
metr(tier 1: the source’s own channel or an official record) - Cite this
- METR, “Review of the "Misalignment in high-stakes settings" section in Anthropic's Risk Report (August 2026)”, October 20, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/metr-2026-10-20-review-anthropic-misalignment-risk-report