US Government Orders Anthropic to Shut Down Fable 5 and Mythos 5 AI Models The world’s most capable public AI model went dark just three days after launch, as U.S. national security officials, Anthropic, big tech rivals, and foreign governments collided over how dangerous frontier systems have become.

On June 9, Anthropic released Fable 5, a commercial spin on its tightly controlled Mythos 5 model, quickly topping benchmarks and “crush[ing] OpenAI’s GPT 5.5 on coding benchmarks by double-digit margins.” Days earlier, CEO Dario Amodei had warned Mythos posed serious national security risks.

By June 12, alarms were ringing in Washington. Amazon shared research with the White House claiming it could “jailbreak and access portions of Anthropic’s powerful new Mythos model that pose a national security threat,” triggering a scramble among senior officials. Commerce Secretary Howard Lutnick sent Anthropic an export‑control letter blocking Mythos 5 and Fable 5 for any foreign person, inside or outside the U.S., a move Axios described as “an escalation in Washington’s effort to treat cutting-edge AI systems as national security assets.”

Anthropic responded that evening by disabling both models for all customers worldwide, saying the government had provided only “verbal evidence of a potential narrow, non-universal jailbreak” and insisting similar vulnerabilities exist “in other publicly available models.” Its public statement stressed it had found only “previously known, minor vulnerabilities” and no “universal jailbreak,” arguing that applying this standard across the industry would “essentially halt all new model deployments for all frontier model providers.”

Fallout spread fast. India, Anthropic’s second‑largest market, saw enterprise deals thrown “into limbo,” supercharging calls for sovereign AI and a multibillion‑dollar domestic fund. Canadian Prime Minister Mark Carney compared the shock to the 2008 financial crisis, warning of “over-reliance on certain models” and urging diversification to avoid systemic AI risk.

In the U.S., cybersecurity leaders argued the shutdown “has taken the best models away from defenders, created market uncertainty, and risked America’s AI leadership without any real risk to justify it,” urging the Trump administration to reverse course. Meanwhile, reports that a China‑linked group may have accessed Mythos, and that Amazon’s warning helped drive the ban, reinforced the administration’s fears about adversaries distilling U.S. frontier models.

As senior Anthropic staff flew to Washington “to try to fix a dispute that has taken the company’s top models offline,” both sides signaled they want a resolution—yet the episode has become a global test case for how far governments will go to treat leading AI as strategic infrastructure.

Continue reading https://foxvector.com/stories/019ecafe-a5e4-24a4-7196-2279c88aa7d9