Anthropic CEO says the era of just studying AI is over, calls for binding US rules

A new essay by Anthropic CEO Dario Amodei has put him in the spotlight after he said the risks of artificial intelligence can no longer be something policymakers just study. He argued that the United States should impose binding safety regulations on the most advanced AI systems.

What happened?

Amodei's remarks are landing at a moment when AI models are becoming more powerful — and more contested. As Decrypt reported, the Anthropic CEO is calling for a Federal Aviation Administration-style framework that would require emerging AI models to undergo mandatory third-party testing before release, with the government able to halt deployments deemed unsafe.

In his new essay, "Policy on the AI Exponential," Amodei says transparency requirements for AI companies are no longer enough. 

"Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast — much faster than the policy process was built to handle. The essay lays out where I think the technology is now, and the action needed to close the gap," Amodei wrote in a post on the social platform X to promote the essay. 

Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast—much faster than the policy process was built to handle. The essay lays out where I think the technology is now, and the action needed to close the gap: https://t.co/Lh6PWae178

— Dario Amodei (@DarioAmodei) June 10, 2026

The policy push comes as Anthropic broadens access to newer models, including a public-facing version of Claude Mythos 5 that automatically redirects certain requests involving cybersecurity, biology, chemistry, and AI development to the less-capable Claude Opus 4.8 as a safeguard against potential misuse.

Decrypt reported that researchers involved with the U.K.'s AI Security Institute found that the Claude model could independently carry out complex cyberattacks, adding to concerns about how these systems are released and supervised.

Why does it matter?

Amodei argued that governments need to prepare for the economic disruption AI could bring — including potential job displacement — while also establishing safeguards against high-risk applications such as cyberattacks, mass surveillance, and the use of autonomous weapons by domestic law enforcement, Decrypt reported.

The proposal would also require AI companies to conduct rigorous safety testing, disclose serious incidents, and prove that their systems meet strict safety benchmarks before they can be released to the public.

Amodei described the issue in stark terms, writing: "Frontier AI models, like airplanes, should be required to go through technical testing and auditing, and their release should be blocked or reversed as a threat to public safety if they do not meet high standards of safety."

Get TCD's free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.