Claude Fable 5 Back Online as US Lifts Export Controls
2 min readAnthropic’s most powerful AI models are back. On July 1, 2026, the US Department of Commerce lifted the export controls that had forced Claude Fable 5 and its higher-tier sibling Mythos 5 offline for more than two weeks, ending a tense standoff between the company and the Trump administration.
How Fable 5 Went Dark
The trouble began on June 12, when Commerce Secretary Howard Lutnick directed Anthropic to suspend access to the two models for all foreign nationals, including the company’s own non-US-citizen employees. Citing national security authorities, the order gave no detailed public rationale. Anthropic received the directive that evening and complied within hours, pulling Fable 5 and Mythos 5 from global access and starting an 18-day shutdown.
Reporting since then points to a specific worry. Officials had reportedly learned of a jailbreak technique that could bypass Fable 5’s safeguards and, by extension, expose the more advanced capabilities built into Mythos 5. The concern was less about who could use the model and more about what a bad actor might unlock inside it.
Claude Fable 5 Returns With New Guardrails
According to CNBC, Lutnick informed Anthropic co-founder Tom Brown by letter that the restrictions were being lifted, crediting the company’s close cooperation with the government. Anthropic did not simply flip the models back on. It redeployed Fable 5 with a new set of classifiers designed to detect and block a wider range of cybersecurity tasks, and said it worked with the US government, Amazon, and other partners to harden the model before restoring access on the Claude platform, Claude.ai, and Claude Code.
Why It Matters
The episode is a preview of a new normal for frontier AI. For the first time, the US government reached in and switched off a commercial model over security concerns, then negotiated the terms of its return. It lands the same week OpenAI’s GPT-5.6 debuted under government-gated access and world leaders gathered in Geneva to discuss AI oversight.
For Anthropic, the quick resolution is a relief, but the precedent is bigger than one company. Frontier models are now treated as strategic assets subject to the same export machinery as advanced chips. Expect more of these interventions, not fewer, as capabilities keep climbing.