Reviews · JUNE 19, 2026
Anthropic pulls Fable 5 and Mythos 5 worldwide after Commerce Department export order
A Friday-evening export-control letter forced Anthropic to disable its two most capable models for every customer on the planet. A week in, the company says access returns 'in coming days' — and reporting points to an Amazon-authored jailbreak paper and an SK Telecom partnership as the trigger.
At 5:21 p.m. ET on a Friday, a Commerce Department directive landed in Anthropic's inbox. By June 12, Fable 5 and Mythos 5 were dark for every customer on the planet. The order didn't specify the security concern in writing, didn't require court approval, and didn't touch Claude Opus 4.8 or any lower-tier model in the stack. It just turned off the two most capable products at a company that had filed confidentially for an IPO earlier that month at a $965 billion valuation.
The stated trigger is a jailbreak paper. Anthropic, in its own statement, describes it as "a potential narrow, non-universal jailbreak, which essentially consists of asking the model to read a specific codebase and fix any software flaws," and says the demonstration surfaced "a small number of previously known, minor vulnerabilities." The Wall Street Journal identified the paper's authors as Amazon security researchers. Cybersecurity Dive reports that Andy Jassy, the Amazon CEO, "warned Trump administration officials that the guardrails were flawed."
The optics are doing a lot of work here. Amazon is Anthropic's largest investor and also the author of the document that produced an emergency export action against the products it competes with through that investment. Anthropic privately asked Katie Moussouris of Luta Security to review the paper. Her read: the gap between "review code for security issues" and "fix this code" is "largely semantic," and the finding "should never have triggered an export control."
By Sunday, 76 signatories had written to Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross arguing the action "has taken the best models away from defenders, created market uncertainty, and risked America's AI leadership." The letter notes that China's frontier models are "only months behind the best American models." It warns: "If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers."
Then there's the Korea piece. The Washington Post, per Korea JoongAng Daily, named a Korean telecom with Mythos access as the company whose suspected China ties prompted the directive. Anthropic counts Samsung Electronics, SK hynix, and SK Telecom among the roughly 150 partners in its Project Glasswing program. At a Seoul press conference this week, Anthropic's Managing Director of International, Chris Ciauri, told reporters: "We are very confident that in the coming days, the models will become available again."
That confidence reads as procedural. The Pentagon already designated Anthropic a "supply chain risk" in March, a finding the company is challenging in federal court. What's new is the mechanism. An export-control letter, sent on a Friday evening with no written rationale, can now reach past a frontier lab's contracts and into its deployment console. The jailbreak is the pretext. The precedent is the point.
Sources
- Statement on the US government directive to suspend access to Fable 5 and Mythos 5 (Anthropic)
- The US government's Anthropic models ban was never about an AI jailbreak (TechCrunch)
- Cybersecurity experts blast US government for restricting Anthropic's AI models (Cybersecurity Dive)
- Anthropic disables Fable and Mythos AI models following U.S. government export ban (Fortune)
- Anthropic says Mythos and Fable 5 access could return in coming days after White House block (Korea JoongAng Daily)