Anthropic Says It's Taking Claude Fable 5 Offline to Comply With US Government Order

Overview

Anthropic says it is taking its Claude Fable 5 AI model offline to comply with an order from the US government. The order asked the company to suspend access for any foreign national, but Anthropic removed access for all customers to ensure compliance. The company says the government cited a national security concern tied to a possible jailbreak, though Anthropic argues the bypass it reviewed was narrow and would not make an attacker meaningfully more dangerous than other available models.

Key Takeaways

Anthropic is taking Claude Fable 5 offline to comply with a US government order it received on Friday at 5:21 pm ET.
The order asked Anthropic to suspend access for any foreign national, but the company removed access for all customers to ensure compliance.
Anthropic says the government believes someone found a method of jailbreaking Fable 5, but the letter did not provide specific details of the national security concern.
The reviewed jailbreak essentially involved asking the model to read a codebase and fix software flaws, and surfaced only known, minor vulnerabilities.
Claude Fable 5, publicly released Tuesday, is a version of Anthropic's Mythos model with safeguards blocking questions about cybersecurity, biology, and chemistry.
The incident is the latest clash between Anthropic and the Trump administration, which earlier labeled the company a 'supply chain risk.'

Stats & Key Facts

#Anthropic received the government letter on Friday at 5:21 pm ET.
#Claude Fable 5 was publicly released on Tuesday.
#The Mythos Preview model had a limited rollout in April.

What Anthropic Did

The company pulled the model to comply with the order.

›Anthropic says it is taking Claude Fable 5 offline to comply with a US government order.
›The order asked it to suspend access to any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
›Rather than limit the block, Anthropic removed access for all of its customers to ensure compliance.

The Government's Stated Concern

The letter cited national security but lacked detail.

›Anthropic received the letter on Friday at 5:21 pm ET.
›The letter did not provide specific details of its national security concern.
›Anthropic's understanding is that the government believes it became aware of a method of bypassing, or jailbreaking, Fable 5.

Anthropic's Assessment of the Jailbreak

The company says the bypass it reviewed was narrow.

›Anthropic reviewed a demonstration used to identify a small number of previously known, minor vulnerabilities.
›These vulnerabilities appeared relatively simple, and other publicly available models could discover them without a bypass.
›The company called the jailbreak narrow and non-universal, saying it essentially consists of asking the model to read a specific codebase and fix any software flaws.

Anthropic said that, to date, the government has only given it verbal evidence of the potential jailbreak, and that one potential jailbreak was shared with the government. The company argued the bypass would not have made an attacker meaningfully more dangerous than they would have been with another AI model.

What Claude Fable 5 Is

The model is a safeguarded version of Anthropic's Mythos system.

›Anthropic publicly released Claude Fable 5 on Tuesday.
›It is a version of the Mythos AI model with safeguards that prevent it from answering questions about cybersecurity, biology, and chemistry.
›Before the public release, the Mythos Preview model had a limited rollout in April.

The limited rollout aimed to give companies and organizations a chance to use the model's cybersecurity capabilities to improve their defenses, while addressing concerns that the technology could be exploited by bad actors to build hacking tools. Anthropic said the public release was conducted in collaboration with the US government.

Tensions With the Trump Administration

This is the latest flashpoint between the two sides.

›Earlier this year, Trump's Department of Defense labeled Anthropic a 'supply chain risk' after the company sought to draw red lines over how the US military could use its technology.
›The designation effectively barred government agencies and contractors from using Anthropic's technology.
›Anthropic responded by filing lawsuits against the Trump administration.

Spokespeople for the White House and US Commerce Department did not immediately respond to WIRED's request for comment.

Anthropic's Position on Process

The company says the order skips the process it supports.

›CEO Dario Amodei said in a policy essay earlier this week that the company supports a fair, structured, and transparent government process to block unsafe AI models.
›In its Friday blog post, Anthropic argued this action does not adhere to those principles.
›Anthropic said it has implemented strong safeguards to reduce the likelihood of Claude Fable 5's misuse.

Anthropic's public stance pairs compliance with criticism: it pulled the model as ordered while arguing the demand fell short of the kind of structured, transparent process Amodei said the company supports. The company also emphasized that the safeguards built into Fable 5 were meant to limit misuse before any government action.

What Is Still Unclear

Several details of the order remain unresolved.

›The letter did not provide specific details of the government's national security concern.
›To date, Anthropic says the government has only given verbal evidence of the potential jailbreak.
›Spokespeople for the White House and US Commerce Department did not immediately respond to WIRED's request for comment.

Anthropic describes the incident as unprecedented and the latest flashpoint with the Trump administration. With only verbal evidence shared so far and limited detail in the letter, the company says its understanding of the concern rests on a demonstration it reviewed rather than a full written explanation from the government.

Frequently Asked Questions

Why is Anthropic taking Claude Fable 5 offline?

It is complying with a US government order that asked it to suspend access for foreign nationals; Anthropic removed access for all customers to ensure compliance.

What concern did the government raise?

The letter cited a national security concern without specific details, but Anthropic understands the government believes someone found a way to jailbreak Fable 5.

How serious does Anthropic say the jailbreak is?

Anthropic calls it narrow and non-universal, saying it surfaced only known, minor vulnerabilities that other public models could also find without a bypass.

What is Claude Fable 5?

It is a version of Anthropic's Mythos AI model, released Tuesday, with safeguards that block questions about cybersecurity, biology, and chemistry.

How does this fit Anthropic's relationship with the government?

It is the latest clash; earlier this year the Department of Defense labeled Anthropic a 'supply chain risk,' and Anthropic sued the Trump administration in response.

Anthropic pulled Claude Fable 5 to comply with the order while publicly disputing both the severity of the cited jailbreak and the process behind the demand.