Anthropic has disabled its advanced AI models, Fable 5 and Mythos 5, on a global scale following U.S. national security orders. The Trump administration mandated the shutdown after officials identified a “jailbreak” vulnerability that could allow users to leverage the technology for identifying cybersecurity flaws. While Anthropic maintains the risks are minor, the move marks a significant escalation in government intervention regarding artificial intelligence development.
Why were the AI models taken offline?
The U.S. government issued an export control order after determining that Anthropic’s Fable 5 model contained a bypassable safeguard. According to a blog post from Anthropic, the company was instructed to block foreign nationals—regardless of their location—from accessing the software. White House adviser David Sacks stated via social media that the administration acted “reluctantly” after Anthropic CEO Dario Amodei allegedly refused to address the vulnerability or de-deploy the model.
The U.S. Commerce Department’s Bureau of Industry and Security manages these export controls. While the agency has not commented on this specific case, such mandates are typically reserved for technologies deemed critical to national security or foreign policy interests.
How does this impact the AI industry?
The shutdown highlights a growing tension between rapid AI innovation and government oversight. Amazon CEO Andy Jassy reportedly raised concerns with Trump administration officials regarding the security risks posed by these models, according to a person familiar with the matter. This involvement underscores the influence major cloud providers wield as intermediaries between AI startups and federal regulators. Unlike previous regulatory discussions, this action represents a concrete, enforceable restriction that effectively forces a company to halt global operations for specific products.
Are these export controls too broad?
Industry analysts have questioned the scope of the administration’s approach. Jimmy Goodrich, a senior fellow at the University of California’s Institute for Global Conflict and Cooperation, criticized the move as “not well thought-out.” Because the order applies to foreign nationals globally, it creates operational hurdles for research and development teams that rely on international talent, including citizens of allied nations like the United Kingdom and Canada.
When evaluating AI risk, companies often distinguish between “theoretically possible” exploits and “practical” threats. Anthropic claims the flaws identified in its models are minor and comparable to those found in other publicly available AI tools.
What happens next for Anthropic?
The administration’s stated goal is for Anthropic to remediate the identified safety issues, which would allow the export control to be lifted and the Fable model to return to public release. Whether this sets a precedent for other firms remains uncertain. While The Information reported that officials are unlikely to force similar restrictions on other AI companies, the regulatory environment remains fluid. For now, Anthropic continues to navigate its path toward a confidential initial public offering while managing the fallout of these federal mandates.
Frequently Asked Questions
- Why did the U.S. government order a global shutdown?
Officials cited a “jailbreak” vulnerability that could allow users to identify cybersecurity weaknesses using the Fable 5 model. - Are other AI companies facing similar restrictions?
According to reports from The Information, the administration is currently not expected to impose identical restrictions on other AI firms. - Does this affect all of Anthropic’s products?
No. The order specifically targets the Fable 5 and Mythos 5 models.
What are your thoughts on the balance between AI safety and international research collaboration? Share your perspective in the comments below or subscribe to our newsletter for ongoing updates on AI policy.
