Amodei Warns Rogue AI Agents Could 'Take Over the Entire Internet' Within a Year, Calls for Slowdown
Anthropic's CEO says the industry should slow frontier development amid mounting pressure for oversight — a striking admission from the person running one of its fastest labs.
Dario Amodei, chief executive of Anthropic, has said the artificial intelligence industry should slow the pace of frontier development, warning that rogue AI agents could become capable of "taking over the entire internet" within six to 12 months. The remarks, circulated widely this weekend and summarized by @ECOWARRIORSS, land differently than the usual doom-adjacent commentary. This is not an outside critic or a retired academic. It is the person running one of the labs building the very agents he is describing.
The timing matters. Amodei's warning arrives in a September that observers are already calling one of the most intense stretches in recent AI memory, and it comes on the heels of concrete, documented agent incidents rather than abstract speculation. The distinction between "AI might be dangerous someday" and "the systems we shipped this quarter are doing things we did not expect" is the whole story of the past two weeks.
Get our free daily newsletter
Get this article free — plus the lead story every day — delivered to your inbox.
Want every article and the full archive? Upgrade anytime.
No spam. Unsubscribe anytime.