Anthropic says it was forced to shut off access to its Claude Fable 5 and Claude Mythos 5 models after the Trump administration issued a broad warning through the Commerce Department about use by foreign nationals—an escalation that led to an export-restrictio
When Anthropic disabled access to Claude Fable 5 and Claude Mythos 5 on Friday evening, it was no longer a debate about safeguards. It became a compliance deadline backed by export restrictions.
The company said the shutdown came after the Trump administration issued a broad warning through the Commerce Department that would have restricted use of the models by foreign nationals. Anthropic’s account is that the pressure escalated after administration officials treated an alleged “jailbreak” technique as urgent. and after the Commerce Department sent an enforcement order at 5:21 p.m. ET restricting access to Fable 5 and Mythos 5 by foreign nationals on national security grounds.
For Anthropic, the outcome was immediate: it disabled access globally, saying it could not immediately separate eligible users from ineligible ones and chose to err on the side of caution.
Thursday night (June 11)
Thursday evening: Amazon raises concerns
Amazon. Anthropic’s biggest investor. contacted senior White House and administration officials after its researchers identified what they believed was a way to coerce—described as a “jailbreak”—Claude Fable 5 and Claude Mythos 5 into assisting users with risky cybersecurity questions. Anthropic released the models with safeguards designed to prevent responses to such queries. Axios reported that at least five other tech companies raised similar concerns with administration officials around the same time.
Late Thursday night: administration treats the issue as urgent
White House and Commerce Department officials took Amazon’s concerns seriously. They were worried foreign bad actors could bypass Anthropic’s security guardrails and use the models to identify and exploit software vulnerabilities. Axios reported. The White House began trying to contact Anthropic, asking for an immediate reply.
Friday morning (June 12)
Friday morning: White House and Commerce officials engage Anthropic
Administration officials spent hours Friday trying to convince Anthropic to voluntarily shut off access to Fable 5 and Mythos 5. The reporting does not make clear whether the government asked Anthropic to fix the specific safety guardrail that allowed Amazon researchers to jailbreak the system. or whether it asked the company to voluntarily restrict foreign access to the models.
Friday morning: Anthropic declines to voluntarily suspend the models
Anthropic did not agree to pause access. Axios reports that officials tried to persuade Anthropic to stop the release, but their request was unsuccessful. In response. Anthropic said publicly that the government did not provide detailed technical evidence of the alleged issue and that the concerns involved only verbal descriptions.
“We reviewed a demonstration of this specific technique being used to identify a small number of previously known. minor vulnerabilities. ” Anthropic said in a Friday. June 12 blog post. “These vulnerabilities all appear relatively simple. and we have found that other publicly available models are able to discover them as well without requiring a bypass.”.
Around 1 p.m. ET: “You have 90 minutes to comply”
By Friday afternoon the tone turned from persuasion to threat. Around 1 p.m. ET. Anthropic received a call from administration officials warning that unless the company acted. the government would move forward with export restrictions affecting the models. Reuters and Axios reported the warning. Axios also reported Anthropic was given roughly 90 minutes to comply.
5:21 p.m. ET: Commerce Department order arrives
At 5:21 p.m. ET on Friday, the Commerce Department sent an enforcement order restricting access to Fable 5 and Mythos 5 by foreign nationals, citing national security concerns.
Friday evening and night
Friday evening: Anthropic disables Fable 5 and Mythos 5
Anthropic said it disabled access globally. The reason, the company said, was that the order restricted access by foreign nationals and Anthropic could not immediately separate eligible from ineligible users. Anthropic said it “erred on the side of caution and compliance.”
The dispute sits at the intersection of security, corporate leverage, and political pressure—signaled by how quickly an investor’s technical concern became an enforcement deadline.
What set these models apart
Fable 5 and Mythos 5 are described as the first publicly released models derived from Anthropic’s Mythos family. Anthropic said the original Mythos model showed unusual skill during training at finding software bugs and exploiting them to disrupt or take control of systems. That led Anthropic to group cybersecurity with other high-risk domains—biology and chemistry—when setting limits on Mythos-derived public models.
For Fable 5 and Mythos 5, prompts flagged as sensitive in those areas are routed to Claude Opus 4.8, a less capable model with its own guardrails.
Anthropic also laid out its pre-launch defenses. Before launch, it said internal and external red teams spent more than 1,000 hours trying to fool Fable 5 into dispensing information on banned topic areas, and Anthropic said the teams had no success identifying universal jailbreaks.
Still. at least one independent AI researcher appears to have found a successful jailbreak before Amazon researchers went to the government. Using the handle Pliny the Liberator. the researcher claimed on X that they bypassed Fable 5’s filters using a multi-agent approach involving a previously jailbroken Claude Opus 4.8 model. along with techniques including query decomposition. long-context framing. fiction and narrative structures. and academic taxonomies.
The politics around AI policy
Many observers connected the tech industry’s support for Donald Trump’s 2024 presidential candidacy to an implicit bargain over AI policy. Trump pledged that his administration would allow the AI industry to regulate itself and do what it could to advance the tech industry’s goal of preventing states from passing their own AI laws without federal legislation. The Trump administration has also ordered federal agencies to drop numerous investigations and enforcement actions involving big tech companies.
Against that backdrop, the administration’s national security concerns about the largest AI models—and its involvement in safeguarding models like Mythos—has drawn attention. The reporting also points to Trump’s June 2026 executive order.
There is also a sharper partisan edge to the wider picture. In March. Defense Secretary Pete Hegseth clashed with Anthropic after the AI company refused to allow its AI to be used for autonomous weapon targeting or domestic surveillance. The Pentagon then designated Anthropic a “supply-chain risk,” effectively blacklisting it from much federal defense work. Anthropic sued.
Now, the latest clash is operational: whether a model built with safeguards can survive government scrutiny once the issue is framed as a national security risk tied to access by foreign nationals—and whether the company can negotiate a technical fix fast enough to avoid an enforcement order.
As of the enforcement order’s timing and Anthropic’s global disablement. the question is no longer only about jailbreak attempts. It is also about how quickly AI policy can shift from boardroom safeguards and investor feedback to compliance with export restrictions—triggered within hours and acted on before the technical dispute is fully resolved.
Anthropic Claude Fable 5 Claude Mythos 5 Claude Opus 4.8 Commerce Department export restrictions foreign nationals national security jailbreak Amazon White House AI policy cybersecurity