Anthropic restores access to Fable and Mythos AI models after US lifts export ban

Anthropic has restored access to Fable 5, its consumer-facing version of its Mythos-class AI model, after safety concerns triggered the US government to impose an export ban. Washington initially ordered Anthropic to restrict access to Fable 5 and Mythos 5, three days after their 9 June launch, banning use by foreign nationals due to fears of devastating cyberattacks. The action followed an Amazon report revealing a method to bypass Fable 5's safeguards to exploit software vulnerabilities.

US Commerce Secretary Howard Lutnick confirmed export controls were lifted after a collaborative analysis. While Mythos 5 was partially released earlier to trusted organisations for defensive cybersecurity, Anthropic agreed to proactively address risks, monitor malicious activity, and establish dedicated teams to work alongside the US government on shared priorities.

The move comes as the US government holds advanced talks with AI developers (£) to create voluntary standards for releasing new models. Expected to be announced next week, these standards will establish benchmarks and timelines for models with advanced cyber capabilities. The initiative implements plans from President Donald Trump's recent executive order on AI, which aims to provide regulatory certainty ahead of anticipated public listings by OpenAI and Anthropic.

In a further development, Sam Altman, chief executive of OpenAI, discussed the importance of establishing a global framework (£) for releasing advanced AI models. The framework should include safety testing and standards, as well as a means for people worldwide to access the benefits that this technology can offer. Altman proposes "a US-led international forum that establishes accepted standards, provides expert and impartial analysis of capabilities and risks, and makes the technology available to nations and companies that participate and follow the rules."

Meanwhile, ahead of its forthcoming initial public offering (IPO), Altman proposes giving a 5% stake (£) to the US government.

In the UK, however, reports have emerged that Andy Burnham is considering an overhaul to the UK's AI strategy (£), shifting away from a US-centric model and instead prioritising British businesses and workers.  

£ - This article requires a subscription.

Training Announcement: Freevacy offers a range of independently recognised professional AI governance qualifications and AI Literacy short courses that enable specialist teams to implement robust oversight, benchmark AI governance maturity, and establish a responsible-by-design approach across the entire AI lifecycle. Find out more.