A frantic 24-hour effort inside the Trump White House ended with export controls imposed on Anthropic, forcing the company to pull its newly released Fable 5 and Mythos 5 models. Senior officials pressed Anthropic CEO Dario Amodei to voluntarily remove the sys
By the time the White House moved, Anthropic’s new model was already in the hands of the public.
Within days of Fable being released. the Trump administration imposed sweeping export controls that forced Anthropic to pull the Fable 5 and Mythos 5 models. The decision came after a frantic 24-hour push by senior officials to convince Anthropic to voluntarily withdraw a newly released artificial intelligence system they believed posed security risks—an effort shaped by tense calls that reached the highest levels of the White House.
The details of those conversations have not been previously reported. Two administration officials and a senior White House official, granted anonymity because they were not authorized to discuss the episode, described the sequence of events and the pressure that built from one call to the next.
The immediate trigger was uncertainty about whether Anthropic’s safety protections could be bypassed. Two days after Fable’s public release. on Thursday. Amazon CEO Andy Jassy raised concerns to the White House about the ability to bypass the model’s guardrails. according to the same two administration officials and the senior White House official. Amazon. an investor in Anthropic. was responding to an administration request for feedback. according to a person familiar with Amazon’s discussions.
By Friday morning, the issue had reached the highest levels of the White House.
In a meeting that included Chief of Staff Susie Wiles. the export-control decision was discussed alongside the models and the administration’s response. according to an administration official and the senior White House official. Treasury Secretary Scott Bessent joined remotely while traveling to Houston for a previously scheduled public event, one of them said.
After the meeting. the administration attempted to reach Anthropic CEO Dario Amodei but was told he was unavailable because he was attending a wellness retreat. according to an administration official and the senior White House official. A spokesperson for Anthropic rejected that claim. saying. “This is absolutely false.” A person close to Anthropic said Amodei was first requested around noon and was on the phone with senior officials within an hour and 15 minutes. While he was out of pocket, Anthropic offered other senior leaders in his place, the person said.
When the administration finally reached Amodei. he took part in three calls with a combination of roughly half a dozen senior administration officials. including White House Cyber Director Sean Cairncross. Bessent. and Commerce Secretary Howard Lutnick. according to the senior White House official and one of the administration officials. Other White House staff and administration officials also participated in some calls: Undersecretary of Commerce for Industry and Security Jeffrey Kessler. White House staff secretary Will Scharf. White House deputy chief of staff Richard Walters. and assistant to the president for policy Walker Barrett. according to the senior White House official.
During the calls, Amodei tried to address what he assumed was a misunderstanding. He pushed back on the administration’s concerns. defended the guardrails. and argued that the specific type of bypass officials described did not pose the same risk as a broader “jailbreak” that would allow the model to be used without any guardrails Anthropic had put in place.
After export controls were put in place. Anthropic said in a blog post that “no testers have yet been able to find a universal jailbreak — a jailbreak method that can very broadly bypass the model’s safeguards. unblocking a wide range of cyber capabilities. ” and that total avoidance of any jailbreaks isn’t now possible for them or any other companies. The company also defended its systems. saying they “are so strong that many users have complained that they are overly broad.”.
Cairncross and Bessent were not persuaded by Amodei’s arguments. A White House official said Amazon’s findings were run past the National Security Agency, and officials felt they had “proof.”
The administration urged Anthropic to voluntarily remove the model and coordinate with the government to address the vulnerabilities. according to the senior White House official and the two administration officials. Amodei asked for more time and information and made no commitments to pull the model, the officials said. At one point. Bessent told Amodei directly that he was making a “bad decision. ” according to the senior White House official.
Shortly after the call(s). the Trump administration imposed export controls on the Fable 5 and Mythos 5 models. citing national security authority and banning their use by foreign nationals. according to Anthropic. The company said the “net effect” of the order was to “abruptly disable” the models for all customers “to ensure compliance.”.
The senior White House official described the export controls as a last step: “Export controls were a last resort after begging them for hours to work with us. This was not something we wanted to do, but our hands were tied.”
But a person close to Anthropic disputed the administration’s characterization that the company was given a genuine choice to voluntarily work with officials. “The White House gave 90 minutes to take the models down, with no details on the actual threat,” the person said. “There was never any begging — or asking — for them to work with us, just a declared 90 minute deadline.”.
White House officials were also taken aback, the senior White House official said, after hearing Amodei liken the dangers of Anthropic’s technology to a nuclear bomb, while still resisting a shutdown aimed at addressing a known security vulnerability.
Anthropic, which has positioned itself as a vocal advocate for AI regulation, has argued that AI poses massive global security risks and could trigger job disruption as the technology advances.
Three people familiar with the government’s thinking said Amazon wasn’t the only company raising concerns. One of them described the problem as an attitude toward the warning itself: “The crux of the issue was the lack of seriousness that Anthropic was applying to it. Had Anthropic taken it seriously and. rather than dismissing it as isolated. moved to fix or pause access. this would have never happened.”.
A second person close to Anthropic refuted the idea that the “jailbreak” was a breakdown of Fable 5’s safety systems and pointed to the company’s collaboration with the administration before it released Fable. The person said the government didn’t object to Fable’s release in multiple conversations.
In its blog post after the export controls were enacted. Anthropic said it was complying with the government’s directive but called the action disproportionate. “As we have stated publicly. we believe the government should have the ability to block unsafe deployments. as part of a statutory process that is transparent. fair. clear. and grounded in technical facts. This action does not adhere to those principles,” Anthropic said.
A White House official, granted anonymity to speak candidly, said innovation remains the White House’s “number one goal, but we also have to prioritize security as well.”
Amazon declined to provide details of its discussions with the administration. “It’s not uncommon for governments to seek our counsel on potential security risks,” an Amazon spokesperson said. “When they occur, we don’t share the details of these discussions.”
The shutdown didn’t come out of nowhere. Anthropic had announced in early April that its latest powerful model. Mythos. would only be available to a limited set of tech and cyber firms. which could use it to test for vulnerabilities in their software. The company said it needed to limit the release because the model was so powerful it could wreak havoc in the wrong hands.
Mythos’s debut helped kick off a series of meetings between Amodei and senior White House officials. Both sides described those meetings as productive. with conversations that ultimately culminated in a recent executive order that requested companies voluntarily submit their advanced models to the government before deploying them widely.
Fable 5, which launched publicly this week, was described by Anthropic as a “Mythos-class model” with safeguards meant to make it safe for general use. The model underwent reviews by the administration and the United Kingdom’s AI Security Institute, according to the account.
But once the alleged security flaws were disclosed, multiple administration officials felt the model needed to be pulled.
In a post on X on Saturday morning. David Sacks—the former White House AI czar and a staunch opponent of regulation—agreed with the administration’s decision to pursue export controls for Anthropic. Sacks wrote that he did not believe the “jailbreak” was simple or not serious. nor did he believe the export controls were an attempt to exert control over the industry more broadly.
“The Admin’s hope now is that Anthropic remediates the safety issue. the export control is lifted. and Fable goes back into general release. ” Sacks wrote. “The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority.”.
Sacks and other officials in the administration have criticized Anthropic before, accusing it of leftist political bias and fearmongering due to its advocacy for stronger regulation and warnings about mass job disruption.
The disputes had already escalated earlier in the year. Secretary of Defense Pete Hegseth and the Pentagon elevated the administration’s disagreements with Anthropic to an unprecedented level by designating the company a supply chain risk on March 3. over Anthropic’s refusal to allow its AI tools to be used for mass domestic surveillance and in autonomous weapons.
On Saturday, Sacks said the earlier feuds between the administration and Anthropic were separate from the export control decision. “The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court,” he wrote.
The episode now sits at the center of a broader clash over how quickly advanced AI should be regulated and how far governments should go when they believe safeguards can be bypassed—especially when the public rollout is already underway.
Anthropic export controls Fable 5 Mythos 5 Trump administration Scott Bessent Sean Cairncross Howard Lutnick AI regulation national security Amazon Andy Jassy
So basically they made the AI disappear? Idk.
I don’t even get why export controls are needed. Like isn’t this just software? Seems like they’re acting shocked it already spread. Also “sys”?? what does that even mean.
Wait, wasn’t Anthropic supposed to be safe? If the White House pressured them to pull it, that sounds like there was some actual scary capability. But then again, these articles always leave out the “security risks” part so it’s kinda hard to tell what’s going on.
How are they gonna stop it if it was already in the public’s hands??? This feels like another one of those ‘we tried really hard’ stories. I saw people online already talking about Fable 5 like 2 days ago, so the export control thing is kinda pointless. Also the article says “forced” but then it’s also “voluntarily remove,” which is it, Trump? Pick a lane.