Anthropic CEO Calls for Global AI Slowdown to Prevent Internet ‘Swarm’ Threat

Anthropic CEO Dario Amodei. (CHANCE YEH/Getty Images for HubSpot)

Key Points

• Anthropic CEO Dario Amodei published an essay calling for a global, coordinated slowdown in artificial intelligence development.
• Amodei wrote that he worries a swarm of AI agents could be capable of taking over the entire internet in six to 12 months.
• OpenAI CEO Sam Altman told staff his company is open to slowing AI development but worries competitors will not coordinate, Bloomberg reported.

After years of rapid development in artificial intelligence, one of the most prominent voices in AI says the industry needs to hit the brakes.

Dario Amodei, chief executive officer of Anthropic, the AI firm behind Claude, published an essay on Saturday calling for a global, coordinated slowdown in AI development.

“We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote. “Progress will still seem fast, and we must make wise use of the time we gain.”

In the essay, titled “We Must Pace the Frontier,” Amodei wrote that following the hack of Hugging Face —an online hub for AI development—by a rogue “swarm” of AI agents being tested by OpenAI, he has become convinced that more prudence is necessary as he and his competitors develop more powerful AI models.

“Given the accelerating rate of AI capability development, it’s my worry that in 6–12 months such a swarm could be capable of taking over the entire internet,” Amodei wrote.

Amodei isn’t the only executive in AI who believes things may be moving too quickly. On Friday, Bloomberg News reported that OpenAI CEO Sam Altman told employees that the company is open to slowing development of its AI models, but that he worries competitors will not want to coordinate.

In his essay, Amodei suggested prohibiting AI to develop biological weapons and wrote that he was in favor of a government-enforced pacing, or a full-on pause, of AI development. Cooperation with China should be pursued where possible, he added.

Amodei admitted that some of his proposals may be outside the realm of possibility.

“The measures I propose to advance the frontier at a safe pace will not be easy,” the essay concludes, “But I believe we owe it to humanity to try.”

Anthropic and OpenAI in June filed S-1 forms with the Securities Exchange Commission to go public. Their initial public offerings are expected this fall and sometime next year, respectively.

Write to Nick Devor at nicholas.devor@barrons.com

Copyright ©2026 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8