Anthropic’s Mythos 5 is back

The Verge
Anthropic has received government approval to restore limited access to its Mythos 5 cybersecurity model for select trusted organizations.

Summary

Following intense negotiations with the Trump administration, Anthropic has secured permission to redeploy its Mythos 5 model to a limited group of cyber defenders and infrastructure providers. While this represents significant progress, the public-facing model, Fable 5, remains unavailable. The government has granted specific exceptions to export control directives, allowing non-US national staff and partner organizations to access the model under strict safety protocols. Both Anthropic and industry peers like OpenAI are pushing for a more standardized regulatory framework to replace the current case-by-case approval process.

(Source:The Verge)