tech
Anthropic's safety warnings may have just backfired — the government has pulled the plug on its most powerful AI
Anthropic isn't hiding its frustration. "We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people," the company wrote in a blog post.
TL;DR
- The U.S. government ordered Anthropic to shut down access to its AI models Claude Fable 5 and Claude Mythos 5.
- The directive was issued due to national security concerns, specifically a claimed jailbreak of Fable 5.
- Anthropic complied with the order but stated it disagrees with the assessment, calling the jailbreak narrow and non-universal.
- The company argues that the identified capability is already available in other public models like OpenAI's GPT-5.5 and is used by cybersecurity professionals.
- Mythos 5 is Anthropic's most capable model, restricted due to its ability to find security vulnerabilities, and was shared with vetted organizations for defensive cybersecurity work.
- Fable 5 was released with guardrails to make it safer for general public use, and benchmark tests indicated it was the most capable public AI model.
- Anthropic believes that applying this standard would halt new model deployments across the industry and found no evidence of safeguards being successfully bypassed to produce harmful content.