Anthropic Reports U.S. Lifts Export Controls on Claude and Mythos

Anthropic launched Claude Fable 5 on June 9, 2026 as its first publicly available Mythos-class model. Three days later, Commerce Secretary Howard Lutnick directed Anthropic to suspend access for all foreign nationals - anywhere in the world, including Anthropic's own foreign national employees. Anthropic received the directive at 5:21pm ET on June 12; the letter provided no specific national security details. In its official statement, Anthropic said: "Our understanding is that the government believes it has become aware of a method of bypassing, or 'jailbreaking' Fable 5." Anthropic examined the technique and concluded it amounted to asking the model to read a codebase and identify software flaws - a capability it said was "widely available from other models (including OpenAI's GPT-5.5)" and used routinely by security defenders. Anthropic complied with the legal directive while publicly disagreeing that a narrow potential jailbreak warranted recalling a model deployed to hundreds of millions of people.

On June 26, Commerce granted Anthropic permission to restore Mythos 5 access for approximately 100 government-approved companies and federal agencies, per Semafor. On June 30, Axios - citing a US official - reported the administration plans to lift Fable 5 controls as soon as Tuesday night (July 1). Commerce Secretary Lutnick posted on X that his office "worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America's leadership in AI." CNBC first reported the full lifting; WIRED added that Anthropic executives flew to Washington for White House meetings during the standoff. What technical or policy changes Anthropic made to satisfy Commerce's concerns remains undisclosed.

The episode is not isolated. OpenAI's GPT-5.6 was simultaneously released only to a select set of approved customers after a government request, per Axios, suggesting emerging norms around staged access for frontier models. China is moving closer to producing its own Mythos-class competitor, per Reuters reporting on 360 Security, while continuing to release open-weight models. The Trump administration faces an August 2026 deadline under a recent executive order to create standardized benchmarks for evaluating the security risk of new AI models - criteria that will determine how future export control decisions are made and what counts as a disqualifying jailbreak.

Enterprises building on Fable 5 or Mythos 5 should monitor updated API access policies, staged rollout tiers for foreign-national users, and any Commerce Department guidance on qualifying jailbreak risk criteria. The August EO deadline on standardized benchmarks is the next concrete frontier model governance milestone.