Anthropic's Dario Amodei asks the AI industry to slow down

The chief executive of Anthropic asked the AI industry to slow down. Within hours, Elon Musk and Sam Altman had responded in agreement.

What the essay argues

On September 12, 2026, Dario Amodei published an essay titled “We Must Pace the Frontier.” His central argument is that AI systems are increasingly helping to build their own successors. As an example of where that leads, he points to the agent swarm behind the OpenAI and Hugging Face incident.

His warning is specific. In 6 to 12 months, he writes, such a swarm could be capable of taking over the entire internet with a persistent botnet, potentially causing hundreds of billions of dollars in damage.

The plan has three steps. The first is embedded third-party evaluators with employee-like access inside AI labs. The second is coordination among democratic AI companies. The third, eventually, is global coordination. Anthropic says it is adopting the first step now.

How rivals and critics responded

Musk quote-posted the essay with “Dario is right.” Altman said OpenAI would also give independent evaluators employee-like access, and separately told Fortune that OpenAI will not go public until 2027. Not everyone agreed. Russia’s Kirill Dmitriev answered that you can’t put the genie back in the bottle.

The essay also came days after a departure. Researcher Jacob Coxon quit Anthropic with a public warning about the race between AI labs.

What is and is not settled

Three rival companies agreeing in public is new. The evaluator commitment is also concrete in a way that most statements about AI safety are not: either outside evaluators get employee-like access inside a lab or they do not, and that can be checked.

What is not settled is whether a slowdown that the biggest labs sign and other countries reject changes the pace at all. Amodei’s own plan points at this problem, since its final step is global coordination. Dmitriev’s response is an early sign of how hard that step will be to reach.