Claude Mythos 5 partially unblocked: US critical infrastructure allowed, Fable 5 still blocked

After two weeks of total lockdown,Claude Mythos 5 ushered in the first partial lifting of the ban. On June 27, Anthropic announced on X: The U.S. government has notified the company,Mythos 5 can be redeployed to U.S. critical infrastructure defense agencies — about 100 companies and federal agencies. However, Fable 5 is still banned and ordinary users are still unable to use it.

Who can use it? Who can't?

According to a letter from U.S. Commerce Secretary Lutnick on June 26, entities receiving exemptions include three categories: Anthropic’s U.S. entities and their foreign employees, Anthropic’s own foreign employees, and U.S. government civilian agencies and national laboratories. According to CNBC, this coversAbout 100 companies and federal agencies.

But the followingAll are excluded:allClaude Pro, Max, Team and Enterprise subscribers, all API developers, and all non-Annex A listed Glasswing partners. For these groups, Mythos 5 and Fable 5 remain inaccessible - any attempted API requests will return an error.

Why only restore Mythos 5 and not Fable 5?

This reflects an accurate calculation by the U.S. government:Mythos 5 is a professional model for cybersecurity, whose advanced offensive and defensive capabilities can both pose a threat and be used for defense. Prioritizing the use of infrastructure defenders is logically tenable - "use the strongest shield to defend the strongest spear."

Fable 5 is a general public version. Its reinstatement will require broader policy decisions - whether to allow "preferential" access for U.S. citizens without completely revoking the export control directive, and how to ensure that models are not resold or leaked outside the United States. These are more complex issues that are being negotiated before the August 1 deadline for the executive order framework.

critical time node

Two structural hallmarks of Fable 5 recovery are:

  • July 8: Anthropic’s updated privacy policy goes into effect, requiring users to provide government-issued identity verification—possibly a mechanism for “US Citizens First” to expand access
  • August 1: In Trump’s June 2 AI executive order, the NSA, Treasury Department, and CISA set a 60-day deadline for completing the cutting-edge model framework—this is the negotiation path for fully restoring Fable 5.

The new normal: an era of government-controlled AI release

The unblocking mode of Mythos 5 and the government-coordinated preview of GPT-5.6 happened almost simultaneously (June 26-27), together establishing a new cutting-edge AI release model:Pre-release government capability review → Initial access to designated partners → Phased expanded rollout → Government signature confirmation. Whether OpenAI "voluntarily" cooperates or Anthropic is "forced" to implement, the result is the same - the U.S. government now determines who can use cutting-edge AI models and when.

For developers, this means an unprecedented variable:A government order can take the model you're relying on offline within 24 hours. Building multi-vendor backup plans is no longer a "best practice" but a survival necessity. access AI Dash Get the latest AI tool reviews and comparisons.

🔗 Share: Twitter Weibo Copy link

📬 Like this article?

Weekly selected AI tool reviews + practical tutorials, delivered directly to you.

Subscribe to the weekly AI picks →

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top
Tool Picks
1
AI Coding
DeepSeek V4 Pro 0813 in-depth review: 1M context + MoE architecture, flagship performance at an affordable price
8.9
2
AI Coding
Qwen3.8 2.4T A95B in-depth review: 2.4 trillion parameter open source weight, a milestone for Alibaba Tongyi Qianwen
8.5
3
AI Coding
GLM 5.3 in-depth review: Zhipu’s millions of contextual reasoning models, fully upgraded programming agent capabilities
8.4
💻AI Coding 📊AI Productivity 📝AI Writing 🎨AI Image Gen
📬 Weekly AI Picks