The Trump administration imposed export controls on Anthropic's advanced AI models citing national security concerns, forcing the company to disable access globally.
8:45 PM
The U.S. government ordered Anthropic to disable access to its Claude Fable 5 and Mythos 5 artificial intelligence models on Friday, June 13, citing national security authorities. The company received the directive at 5:21 p.m. ET instructing it to suspend all access to the models by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
Unable to differentiate between American and foreign users, Anthropic disabled the models for all customers to ensure compliance with the export control directive. The company said all other Claude models remain available and unaffected by the order.
Fable 5 is the constrained version of Mythos, the AI model Anthropic announced in April. The company had released Fable 5 to the public on June 9, describing it as the most powerful AI model it had ever made widely available. The model came with safeguards restricting responses to certain high-risk queries, particularly in areas including cybersecurity, biology, and chemistry.
According to reporting, Amazon security research contributed to the government's decision. Amazon researchers reportedly discovered that through a series of prompts, Fable 5 could provide information usable in cyberattacks. Amazon CEO Andy Jassy shared the company's findings with the White House, which subsequently issued the export control order.
Anthropic disputed the government's characterization of the issue as a "jailbreak." In a statement, the company said the government "did not provide specific details of its national security concern" and that evidence of potential vulnerabilities was provided only verbally. Anthropic argued that many of the same vulnerabilities could be discovered using other publicly available models, including GPT 5.5, and that the discovered vulnerabilities were minor.
The company outlined steps it had taken to safeguard Fable and Mythos, including working with the U.S. and UK governments and changing its data retention policy to help track attempts to use the models for malicious purposes. Anthropic stated it had not received a full disclosure of the government's specific concerns.
Trump administration officials concluded talks with Anthropic on Monday without lifting the export controls. According to three people briefed on the matter, the administration continues to believe there are ways to disable some of the guardrails on Claude Fable 5, effectively allowing users to access the more powerful cybersecurity capabilities of Mythos. Anthropic has maintained that the administration's concerns are overblown, a position it reiterated in working group meetings held at the Commerce Department with government researchers from the Center for AI Standards and Innovation and the Office of the National Cyber Director.
The incident marks the latest flashpoint between Anthropic and the Trump administration. Earlier this year, the Department of Defense labeled Anthropic a "supply chain risk" after the company sought to draw restrictions on how the U.S. military could use its technology, effectively barring government agencies and contractors from using Anthropic's systems.
The shutdown has also sparked broader industry discussion about the risks of relying on closed AI models. Companies building on proprietary systems face the possibility of access being cut off without warning. Chinese open-source AI companies, including Zhipu and MiniMax, saw investor interest surge as firms considered the advantages of open-source models that can be downloaded, run on private infrastructure, and customized around proprietary data.
Prior to the government order, Anthropic had already faced criticism over Fable's design. The company apologized for implementing hidden guardrails that throttled the model in ways not fully transparent to users and researchers. Anthropic said it would reverse course and be more transparent about when restrictions activate, even if that means refusing more queries. Additionally, Microsoft restricted Claude Fable 5 for internal employee use due to concerns about Anthropic's new data retention requirements, citing potential risks to customer data and confidential information.
The U.S. government ordered Anthropic to block foreign citizens from accessing its Fable 5 and Mythos 5 AI models, citing national security concerns.
Pope Leo XIV released "Magnifica Humanitas," a 42,300-word encyclical Monday addressing artificial intelligence's threat to human dignity and calling for strict regulation and "disarming" of the technology.
Homeland Security Secretary Markwayne Mullin said the administration is drawing up plans to stop processing international flights in cities with sanctuary laws, citing protests at an ICE detention facility in New Jersey.
Meta will lay off approximately 8,000 employees starting May 20 and halt hiring for 6,000 open positions.
A federal court blocked the fund temporarily; the DOJ said it will comply with the ruling despite disagreeing with it.
The ultra-low-cost carrier ceased operations Saturday after rescue talks with the Trump administration collapsed, affecting 17,000 employees and canceling all flights.