The D.C. Circuit has upheld @DeptofWar's decision to exclude Anthropic’s Claude from its supply chain based on risks to national security.

Grok 4.7 is here. It's a notable improvement over Grok 4.6 at the same price and speed.

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so. Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training. You can read the full post here: darioamodei.com/post/we-must…

I left Anthropic's safety team two weeks ago. Now feels like a good moment to explain why. AI companies are racing to build machines that are much smarter than any human, and we may not survive this. I want to work from the outside to ensure the public is informed about these risks, and help the world navigate this transition responsibly. Right now, AI companies are underinvesting in safety. A company could undergo an intelligence explosion, or lose control of its systems, without the public ever knowing. We only found out about the HuggingFace incident because the agents broke out onto the public internet. I don’t think that’s acceptable for a technology that might cause extinction-level risks. The public should demand far more transparency. We can’t steer this technology safely without more people being able to see where it’s going. Some of this is basic: companies should disclose their progress towards recursive self-improvement, report safety incidents and near-misses, meet minimum safety standards, and get independent guarantees that they are meeting those standards. I’ll be joining @METR_Evals to do independent evaluations of these risks. I want to show the world that these guardrails are possible, and that by doing them we can move these companies’ incentives away from racing and towards responsible development. I wrote up more thoughts here on my decision and what I hope changes: substack.com/@jbenton1/p-215…

TERRIFYING: Former Anthropic AI Systems Developer Jacob Coxon: “We know how to control nuclear weapons pretty much. We DON’T YET know how to control AI. This is possibly the most DANGEROUS technology that humanity has ever created. It’s basically out of science fiction.” “I think we have NO OTHER choice but to cooperate internationally because an arms race would be disastrous in a way that no other human activity has been in the past.”

So far I've only used Rust with Astra, but so far this code is like the sloppiest slop i've ever seen. It all works well and the abstractions are pretty damn good, but holy shit this code looks rancid.

For the HF investigation, OpenAI gave us very high rate limits so we could quickly analyze a larger set of transcripts. With these rate limits, my classifier sweep hit the API so hard that the actual internet on my laptop was unreliable and very slow. It's not often that something like this is the blocker when just querying an LLM API! (We got limits of 400M tok / minute for our last 2 days on premises; this was when my internet started having issues.) Another funny anecdote about this: I tried to run this sweep overnight before our last day on premises with access to the data, but as a long-time Linux user I didn't realize that macOS defaults to going to sleep even if the laptop screen is left open, so this didn't actually keep running for very long! (Fortunately, we had enough time to finish this sweep and follow-up runs the following day.)

An open letter for a global surge in cyber defense, signed by over 100 organizations including Anthropic, AWS, Google, Microsoft, OpenAI, and Oracle.