US Lifts AI Export Ban After Anthropic Strengthens Safeguards

One of the world's leading AI companies just got the green light to restore global access to its most advanced technology after proving it could be used safely.

Anthropic announced Tuesday that the US Department of Commerce lifted export restrictions on its Fable 5 and Mythos 5 AI models. The company will begin restoring worldwide access starting Wednesday.

The restrictions came suddenly on June 12 when the government discovered vulnerabilities in the safeguards meant to prevent misuse of the powerful tools. For nearly three weeks, only certain American cybersecurity firms could access the technology while Anthropic worked to address national security concerns.

Secretary of Commerce Howard Lutnick confirmed in a June 26 letter that Anthropic had successfully worked with the US government to address the risks. By Tuesday, the Trump administration was satisfied enough with the improvements to withdraw all previous restrictions.

The collaboration happened under a June 2 executive order calling for a voluntary framework where private AI companies could test and release their most powerful models in partnership with the government. Anthropic became one of the first to navigate this new process successfully.

Rival company OpenAI is taking a similar path, restricting its new GPT-5.6 model to approved partners only. CEO Sam Altman acknowledged the process isn't perfect but appears willing to work within the framework.

The Ripple Effect

This partnership model could reshape how breakthrough technology reaches the world. Rather than companies releasing innovations without oversight or governments blocking them entirely, the collaborative approach lets safety improvements happen while keeping progress moving forward.

White House Chief of Staff Susie Wiles thanked companies for their cooperation on testing and security guardrails. The successful collaboration shows that protecting national security and advancing AI development don't have to be opposing goals.

CIA Director John Ratcliffe compared the most advanced AI models to "digital nuclear weapons" in terms of their capabilities. That serious assessment makes the collaborative testing framework even more important for future releases.

When innovation and security work hand in hand, everyone benefits from safer, more responsible technology.