Blab Tech
Anthropic Faces U.S. Government Standoff Over Claude Fable 5 and Mythos 5
When Washington’s Commerce officials crossed paths with Anthropic’s top brass in the capital, the meeting was more a chess match than a handshake. On Monday, executives from the AI‑lab sat down with representatives from the U.S. Department of Commerce to debate whether Claude Fable 5 and Claude Mythos 5—its newest large‑language models—pose a national‑security risk.
The talks ended without a resolution, leaving both sides firmly entrenched. Anthropic insists the models were fully vetted and approved by the government before their public launch on June 9, 2026. Its only identified issue is a narrow jailbreak that could let a user bypass a cybersecurity guardrail and ask Fable 5 to read a specific codebase to fix bugs. The company argues that recalling a model already in use by hundreds of millions would set a dangerous precedent, potentially stalling future deployments across the industry.
The dispute gained steam after the Pentagon named Anthropic a “supply‑chain risk” in March 2026—a label usually reserved for foreign adversaries. Defense Secretary Pete Hegseth warned that the firm was a threat and barred Claude from Pentagon facilities. On June 12, the U.S. government ordered Anthropic to suspend access for all foreign nationals to Fable 5 and Mythos 5 on national‑security grounds. The next day, the company announced it would “abruptly disable” the models for all users.
David Sacks, former Trump adviser on AI, took to X to criticize Anthropic’s response, saying the firm’s stance was “at odds with its branding and ethos as a safe‑AI research firm.” He added that the administration expects the company to fix the safety issue, lift export controls, and return the models to general release.
Anthropic has framed the disagreement as a misunderstanding and says it is working to restore access as soon as possible. Public statements emphasize that the government’s concern centers on a single narrow jailbreak, not a broader threat.
The standoff is being viewed as a test case for how aggressively Washington will regulate frontier AI models. No timeline has been set for when—or if—Fable 5 and Mythos 5 will return to public access. The models remain unavailable to U.S. users, and Anthropic’s future compliance with federal safety and export‑control requirements remains uncertain.
Founded in 2021 by former OpenAI researchers Daniela and Dario Amodei, Anthropic has positioned itself as a leader in AI safety, developing the Claude series and the Mythos line. Its valuation was estimated at $965 billion in May 2026, making it the most valuable pure‑play AI company.
The current situation reflects a clash between a private AI firm’s safety claims and a government’s national‑security assessment. Anthropic’s next steps will likely involve further technical reviews, potential model modifications, and negotiations with Commerce and Defense officials. Until those discussions conclude, the models remain inaccessible to U.S. users and foreign nationals alike, and the outcome will shape future U.S. policy on advanced AI systems and influence how other companies approach safety and export controls for frontier models.
The talks ended without a resolution, leaving both sides firmly entrenched. Anthropic insists the models were fully vetted and approved by the government before their public launch on June 9, 2026. Its only identified issue is a narrow jailbreak that could let a user bypass a cybersecurity guardrail and ask Fable 5 to read a specific codebase to fix bugs. The company argues that recalling a model already in use by hundreds of millions would set a dangerous precedent, potentially stalling future deployments across the industry.
The dispute gained steam after the Pentagon named Anthropic a “supply‑chain risk” in March 2026—a label usually reserved for foreign adversaries. Defense Secretary Pete Hegseth warned that the firm was a threat and barred Claude from Pentagon facilities. On June 12, the U.S. government ordered Anthropic to suspend access for all foreign nationals to Fable 5 and Mythos 5 on national‑security grounds. The next day, the company announced it would “abruptly disable” the models for all users.
David Sacks, former Trump adviser on AI, took to X to criticize Anthropic’s response, saying the firm’s stance was “at odds with its branding and ethos as a safe‑AI research firm.” He added that the administration expects the company to fix the safety issue, lift export controls, and return the models to general release.
Anthropic has framed the disagreement as a misunderstanding and says it is working to restore access as soon as possible. Public statements emphasize that the government’s concern centers on a single narrow jailbreak, not a broader threat.
The standoff is being viewed as a test case for how aggressively Washington will regulate frontier AI models. No timeline has been set for when—or if—Fable 5 and Mythos 5 will return to public access. The models remain unavailable to U.S. users, and Anthropic’s future compliance with federal safety and export‑control requirements remains uncertain.
Founded in 2021 by former OpenAI researchers Daniela and Dario Amodei, Anthropic has positioned itself as a leader in AI safety, developing the Claude series and the Mythos line. Its valuation was estimated at $965 billion in May 2026, making it the most valuable pure‑play AI company.
The current situation reflects a clash between a private AI firm’s safety claims and a government’s national‑security assessment. Anthropic’s next steps will likely involve further technical reviews, potential model modifications, and negotiations with Commerce and Defense officials. Until those discussions conclude, the models remain inaccessible to U.S. users and foreign nationals alike, and the outcome will shape future U.S. policy on advanced AI systems and influence how other companies approach safety and export controls for frontier models.
