Anthropic switched off public access to two of its most capable AI models — Fable 5 and Mythos 5 — on June 12, 2026, after the U.S. government ordered the company to cut off all foreign access on national security grounds. The shutdown applies to every customer, not just users abroad. All of Anthropic’s other models stayed online.
It’s a first. No major AI lab had been pushed to pull a flagship product under direct government pressure before this.
What the government ordered
According to Anthropic’s own statement, the directive arrived at 5:21 p.m. ET on June 12, citing national security authorities. It required the company to suspend access to Fable 5 and Mythos 5 for “any foreign national” — anywhere in the world, inside or outside the United States.
That scope created a problem. To guarantee compliance, Anthropic said it had to abruptly disable Fable 5 and Mythos 5 for all customers, because it could not selectively block only foreign nationals fast enough. Access to every other model, including Opus 4.8, was not affected.
Why it happened
The trigger, per Anthropic, was a suspected “jailbreak” — a way to bypass the model’s safety guardrails. The company says the government believes it found a method of getting around Fable 5’s protections.
The specific technique, as Anthropic describes it, was narrow: asking the model to read a particular codebase and fix its software flaws. That matters because Fable 5 belongs to Anthropic’s Mythos class. The model had only gone public on June 9, three days before the order, and was built to route sensitive cybersecurity, biology, and chemistry questions to the older Opus 4.8 model.
Anthropic pushed back
Anthropic complied with the order but openly disagreed with it. The company argued that the finding of a narrow potential jailbreak should not be reason to recall a commercial model deployed to hundreds of millions of people, and characterized the episode as a misunderstanding with the government.
The government has not, as of this writing, published a detailed technical justification. This is a developing story, and the situation may change quickly. The directive was also reported by Bloomberg, NBC News, and CNBC.
What it means for everyday users and small businesses
For most people and businesses, the immediate practical impact is small. Anthropic’s other models still work, and competing AI tools are unaffected. But the episode is a useful reminder of a bigger trend: the same AI that can write your code or fix a flaw can also be turned toward finding weaknesses in someone else’s systems.
That’s the real takeaway for a small business. AI capability and cybersecurity risk are now two sides of the same coin. The fundamentals haven’t changed — keep software patched, use strong authentication, and don’t assume any single tool is risk-free. For a plain-English starting point, our team covers the basics in our cybersecurity tips for small businesses and our ransomware guide.
Quick questions
Is Fable 5 banned permanently? No. As of June 12, 2026, access is suspended to comply with a government directive. Anthropic has not announced a timeline, and it disagrees with the order, so the status could change.
Are other AI models affected? No. Anthropic said only Fable 5 and Mythos 5 were disabled. Other models, including Opus 4.8, remain available.
Does this affect my business in Charlotte? Not directly. It’s a national, model-specific action. The broader lesson — AI is now deeply tied to cybersecurity — is what matters most for local businesses planning their tech.
John Jones
Senior IT Specialist, IT Carolina
John has 12 years of hands-on experience diagnosing and resolving computer, printer, and network issues for homeowners and small businesses across Charlotte, NC. He has helped hundreds of clients recover from Windows update failures, driver conflicts, and hardware problems — often resolving in a single remote or on-site session.
Share this article:
