Elon Musk backs Anthropic CEO’s call to slow AI development

Anthropic CEO Dario Amodei on Saturday initiated a new push to deliberately slow the pace of frontier model development in the midst of growing calls for safety.

In an essay titled “We Must Pace the Frontier” Amodei issued a chilling warning that if the tech companies fail to curb unchecked growth of AI models, the safety guardrails and regulations fall short of it, thereby leading to “catastrophic damage.”

As a result, the autonomous AI agents could hack global internet infrastructure within the next year.

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

He also proposed a three-step framework in which Anthropic would install permanent third-party reviewers inside frontier AI models with access to relevant tools and internal risk assessment processes.

Amodei clarified that he wasn't advocating for a freeze on training or technical advancement. Instead, he emphasized the need for companies to dedicate sufficient time to model alignment and safety, with third-party evaluators verifying those safeguards.

Elon Musk also agreed with Anthropic's call to slow the development of AI models. Taking to X, he posted, “Dario is right.”

Anthropic’s essay comes days after when the Anthropic researcher resigned from his post. He also warned the world that the tech companies are gambling with their lives by acting irresponsibly.