US Removes Export Controls on Anthropic’s AI Models After Security Review

The US government has just flipped the switch. After a tense more-than-two-week blackout, export controls on Anthropic’s powerful AI models, Fable 5 and Mythos 5, are lifted. This means foreign nationals can now access these models again. What changed? Security fears that sparked the ban have been addressed head-on with new safeguards and thorough government testing.

From Launch to Lockdown

It all started on June 9, 2026. Anthropic launched Claude Fable 5, a major leap in AI tech. But just three days later, alarm bells rang in Washington. The US government ordered Anthropic to restrict access to both Fable 5 and Mythos 5 models for all foreign nationals. The reason? Fears that these advanced AIs could be hijacked for cyber-attacks by hostile foreign actors.

Anthropic had to pull the plug entirely on access for foreign users. This wasn’t a small hiccup. It was a hard stop. The blackout lasted more than two weeks as the company scrambled to tighten security and prove the models were safe to share again.

Testing and Tougher Safeguards

After the export ban, Anthropic didn’t just sit still. They ramped up safeguards and opened the door to closer cooperation with the US government. Howard Lutnick, the US commerce secretary, confirmed this week: “Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America’s leadership in AI.”

Researchers from the US Department of Commerce’s Center for AI Standards and Innovation put Anthropic’s new protections through the wringer. Their verdict? The safeguards are extraordinarily strong — strong enough to ease the government’s cybersecurity worries and lift the restrictions.

Before the full reopening, the US government allowed Anthropic to release Mythos 5 to a select group of trusted US organizations. This limited rollout kept the AI in secure hands while the final security review went on.

Global Cybersecurity Concerns and Collaboration

The US isn’t alone in sounding the alarm about AI risks. Intelligence agencies from Australia, the UK, New Zealand, and Canada joined the chorus. They warned that AI could accelerate cyber threats dramatically, with a timeline of just months before risks would exceed expectations.

Anthropic reacted by pledging to scale up government collaboration. Dedicated teams will focus on shared priorities between the company and regulators. This move aims to keep AI development transparent and safe, while pushing innovation forward.

Even political figures like Susie Wiles, Donald Trump’s former chief of staff, weighed in. She publicly thanked companies for working with the White House on AI security. This signals heightened political attention on AI risks and partnerships with industry leaders.

What’s Next for Anthropic and AI Security?

The lifting of export controls marks a new chapter. Anthropic’s AI models are back in the global game, but with fresh eyes on safety and regulation. The company’s promise to deepen government ties shows a clear path forward. They want to lead on AI innovation without sacrificing security.

This episode highlights a bigger story. AI is advancing fast and governments are scrambling to keep pace. Cybersecurity fears are real and urgent. But solutions are emerging through tough scrutiny and collaboration. The US decision signals confidence that AI can be both powerful and safe—if handled right.

As AI evolves, so will the battles over access, control, and trust. Anthropic’s journey from a sudden export ban to a green light shines a spotlight on what’s at stake. We’re watching the future unfold, where innovation meets responsibility in real time.