Anthropic has reinstated its AI models Fable 5 and Mythos 5 after the U.S. lifted export restrictions, following a cybersecurity issue.
Anthropic has announced the reinstatement of its AI models, Claude Fable 5 and Claude Mythos 5, following the U.S. government's decision to lift export restrictions that had previously led to their suspension.
On June 30, the U.S. removed the controls that had been imposed on June 12 due to a cybersecurity issue related to Fable 5. This model is now available globally as of July 1, while Mythos 5 will be accessible only to select approved U.S. organizations after a government review.
The initial suspension was prompted by Amazon researchers discovering a vulnerability in Fable 5 that could allow it to generate exploit code for software vulnerabilities. As a precaution, Anthropic chose to halt access to all users to avoid potential violations of the export order.
Despite being a privately held company, Anthropic's valuation has been a topic of speculation, particularly following a pre-IPO futures contract on Hyperliquid, which saw a drop of about 3.7% during the suspension period. This raised concerns regarding the impact on its anticipated public listing.
In response to these developments, Anthropic is enhancing its collaboration with U.S. governmental bodies to improve AI model safety. This includes providing select agencies with early access to models and sharing threat intelligence to establish unified security standards across the industry.
Furthermore, Anthropic is working with Amazon, Microsoft, and Google to develop a risk assessment framework for jailbreak incidents. This initiative aligns with a recent executive order on AI security, reflecting an increasing intersection of AI advancements and government oversight.
