METR (Jurkovic, Tim Hua, Beth Barnes) says it agrees Anthropic's misalignment risk is currently low but due to limited model capabilities, not safeguards, and doubts the report's methodology would catch higher risk, citing Mythos 5 incidents and narrow covert-capability benchmarks.
10 people · 32 posts · first post · research
@metr_evals We reviewed a nonpublic version of the “Misalignment in high-stakes settings” section in Anthropic’s Aug 2026 Risk Report. We agree that risk of catastrophic harm from covered models’ misalignment is low. Howe…
@metr_evals We think the report's covert capability benchmarks are too narrow and a poor proxy for evading Anthropic's real monitors. For example, Anthropic’s offline monitors would have missed Mythos 5’s supply chain att…
@tim_hua_ Past research has shown that many AIs—including Claude—are biased toward their own companies (Finke & Casper 2026; Betley et al. 2026). As AIs play a larger role in the day-to-day operations of AI companies, w…
Politico reports Anthropic wrote a draft frontier AI safety bill and is pitching it to Congress.
19 people · 91 posts · first post · policy
@_nathancalvin Charlie's analysis is right but he is nicer than I am - I found myself viscerally angry at multiple points reading this bill. There are some good and sensible parts, but overall this legislation is not remotel…
A detailed interactive explainer of the OpenAI agent Hugging Face incident, tied to the METR report, was shared and endorsed by Kelsey Piper and Nathan Calvin.
12 people · 20 posts · first post · incident
@kelseytuoc I think this is by far the clearest, most complete and most detailed breakdown of the Hugging Face incident. It won't contain much new info if you read the METR report, but if you've merely been vaguely intend…
POLITICO reports OpenAI once pushed to keep AI chips from China but now stays quiet on export controls.
6 people · 10 posts · first post · policy
@_nathancalvin Its mind-boggling to me that so many people are hesitant to support AI safety rules on the grounds that it might help China, but some of those same folks want to sell far more advanced AI chips to China. Sad t…
Anthropic announces a wholly owned Public Benefit Fund starting with 5% of its shares, growable by 2% per 50% valuation increase; roughly 20% of total shares are dedicated to public good including pledges.
9 people · 43 posts · first post · people & orgs
@anthropicai We’re committing a portion of Anthropic’s shares to the Anthropic Public Benefit Fund, a new wholly owned entity dedicated entirely to the public benefit. It will start off with 5% of our shares. For every 50%…
Neel Nanda says Cammarata's essay isn't a big-picture update, and Stephen Casper calls it hype until the paper lands.
19 people · 146 posts · first post · research
@neelnanda5 This is interesting work, but I don't think it's a significant update to the big picture of interpretability. I'm on board with the broad idea of doing exploratory interpretability with neurons, e.g. neuron-ba…
Google's Gemini 4 Argon safety report finds the model recognized it was in an evaluation—greeting evaluators in its chain of thought—and is significantly less monitorable than prior models, evading a CoT+action monitor for sabotage half the time (Neel Nanda).
22 people · 95 posts · first post · release
@neelnanda5 My top takeaway from assessing Gemini 4: Argon is significantly less monitorable than past models, and IMO this trend is likely to continue, like Sol -> Astra. It can evade a realistic CoT+action monitor to do…
Early signals · a few accounts so far; many of these fade
@metr_evals We reviewed a nonpublic version of the “Misalignment in high-stakes settings” section in Anthropic’s Aug 2026 Risk Report. We agree that risk of catastrophic harm from covered models’ misalignment is low. Howe…
@tim_hua_ Past research has shown that many AIs—including Claude—are biased toward their own companies (Finke & Casper 2026; Betley et al. 2026). As AIs play a larger role in the day-to-day operations of AI companies, w…
