After almost three weeks, the US government has lifted the export controls on Anthropic’s Fable 5 and Mythos 5 AI models. In a post on X, the company said it will begin restoring access tomorrow, while promising to share another update soon. Soon, all foreign nationals will be able to use the firm’s newest AI model.
For context, the company unveiled Fable 5 and Mythos 5 in early June as its latest and most capable AI models. Just days later, the Trump administration ordered the firm to suspend access to both models for all foreign nationals.
In an official statement, Anthropic said the US government issued the directive over concerns that someone could find a way to “jailbreak” Fable 5’s safeguards. However, after reviewing the model, the company said it had identified only a limited number of relatively minor software vulnerabilities, none of which could bypass the model’s safeguards on their own.
Despite Anthropic’s findings, the US government proceeded with the export restrictions on both Fable 5 and Mythos 5. It later granted select companies and government agencies access to the AI models, before announcing a broader restoration of access today.
“Over the past two weeks, we have worked closely with Anthropic to analyse and approve Fable 5 to ensure alignment across the US Government and strengthen America’s leadership in AI,” Commerce Secretary Howard Lutnick wrote on X. Shortly afterwards, the AI company confirmed that the export controls on Fable 5 and Mythos 5 had been lifted, adding that it would begin restoring access to both AI models the following day.
According to a report by the Financial Times, US Commerce Secretary Howard Lutnick has informed Anthropic that the restrictions on its Fable AI model have been lifted after the company committed to proactively identifying and addressing security risks related to the model. The Commerce Department had previously imposed restrictions on Fable and Mythos following concerns raised by Amazon and other companies that the models could be jailbroken and potentially exploited for malicious purposes.
While the Trump administration has lifted the restrictions on Anthropic’s AI models, that doesn’t mean other models are in the clear. Just days earlier, the US government issued another directive, this time targeting OpenAI’s GPT-5.6, ordering the company to limit access to the model to trusted partners only. At the time of writing, it remains unclear when OpenAI will receive permission to make the model publicly available.
(Source: Anthropic [X] / WIRED / Financial Times)
