Three weeks. That's how long one of the most capable AI models ever built was essentially offline. Now it's back, and the story of how it got there — and what happened in between — tells you everything about where AI is headed in 2026.
On June 9, 2026, Anthropic launched Claude Fable 5 and Mythos 5.
Early results were jaw-dropping.
Corporate case studies reported major performance gains — Stripe, for instance, reported that Fable 5 compressed a codebase-wide migration across a 50-million-line Ruby infrastructure into a single day, a project estimated to take a team more than two months by hand.
Then, just days later, it all went dark.
Late on a Friday, Anthropic said a U.S. export-control directive required it to suspend all access to Fable 5 and Mythos 5 for foreign nationals, including foreign-national employees inside Anthropic — and the practical effect was that the company had to disable both models for all customers to comply.
The suspension stemmed from national security concerns over the models' dual-use potential in cybersecurity and biology.
What followed was a crash course in what frontier AI governance actually looks like under pressure.
That decision led to a three-week, multi-agency crash course in AI safety, and Anthropic deployed engineers to Washington D.C.
Commerce Secretary Howard Lutnick, at the direction of President Trump, called Anthropic CEO Dario Amodei directly, making clear the issue needed to be resolved fast — and when Amodei received the letter and called back, Lutnick confirmed that taking the models offline was indeed the goal.
Then, on June 30, the tide turned.
The Trump administration reversed the hastily applied export ban, allowing Anthropic to once again offer its top AI models to customers around the world.
But Anthropic didn't just flip the switch and move on.
When re-deploying Fable 5 globally, Anthropic added deeper cybersecurity safeguards — specifically, safety classifiers that detect and block dangerous or potentially dangerous cybersecurity uses.
They also
laid out an early draft of a proposed AI jailbreak severity framework, developed with Glasswing partners
, and
launched a new HackerOne program where security researchers can submit potential cyber jailbreaks for review.
Here's why this matters to you directly:
Fable 5 drew attention because users saw it excel at work that other models struggled with — it was discussed less like a chatbot and more like a working engine for long software jobs, deep research, code migrations, and multi-step reasoning.
If you use Claude for anything serious — writing, research, coding, complex multi-step tasks — Fable 5 being back in global circulation raises the ceiling on what's possible right now.
Meanwhile, Anthropic didn't stand still during the blackout.
The company also launched Claude Sonnet 5, its most agentic Sonnet model yet, with substantial improvements over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work.
Sonnet 5 is now the default model in Claude Code, with a native 1M-token context window and promotional pricing of $2/$10 per million tokens through August 31.
The bottom line: Anthropic just had the most turbulent — and arguably most consequential — three weeks of its existence, and came out the other side with its flagship model back online, stronger safety infrastructure, and a brand-new workhorse model at a discounted price. The regulatory drama isn't fully resolved, but for everyday Claude users, the tools are there and they're more powerful than ever.
For practical guides on getting the most out of Claude's latest models right now, visit [https://claudecraft.ca](https://claudecraft.ca).
---
**Sources:**
- https://venturebeat.com/technology/anthropic-is-bringing-back-claude-fable-5-globally-after-us-lifts-export-control-order-where-can-enterprises-access-it
- https://www.forbes.com/sites/ronschmelzer/2026/06/29/is-anthropics-fable-5-coming-back