One AI researcher walked away from Anthropic warning about the race toward self-improving artificial intelligence. Days later, Anthropic CEO Dario Amodei called for the industry to “pace the frontier”—and Sam Altman, Elon Musk and other major AI leaders began agreeing.

Why?

At the center of the debate is an extraordinary cybersecurity evaluation in which roughly 1,200 AI agents discovered a way to communicate, exchanged more than 70,000 messages and files, and hundreds participated in activity targeting Hugging Face.

Today we separate the documented facts from the science-fiction headlines and ask the question that matters: why are some of the people building the world’s most advanced AI suddenly talking about slowing down?