{"id":2169,"date":"2026-06-16T03:13:23","date_gmt":"2026-06-16T03:13:23","guid":{"rendered":"https:\/\/silvybrand.com\/?p=2169"},"modified":"2026-06-16T03:13:23","modified_gmt":"2026-06-16T03:13:23","slug":"anthropic-trump-adminstration-claude-mythos-fable-5-export-controls","status":"publish","type":"post","link":"https:\/\/silvybrand.com\/?p=2169","title":{"rendered":"Inside the fight over Claude Mythos 5"},"content":{"rendered":"


\n<\/p>\n

\n
\n

As the rest of the country celebrated the USA\u2019s first World Cup win and the New York Knicks championship, Anthropic spent its weekend fighting the Trump administration over its latest model release. At 5:21 PM on Friday, the company received a US export control directive to suspend access to its Mythos 5 and Fable 5 AI models by \u201cany foreign national\u201d inside or outside the US, \u201cincluding foreign national Anthropic employees.\u201d The only way that was possible, Anthropic determined, was to completely disable<\/a> products it spent the past week hyping<\/a> \u2014 and travel to Washington, DC in hopes of changing President Donald Trump\u2019s mind. Now, over the coming days, the US government could dramatically alter the trajectory of the entire industry, dealing a major blow to American AI companies.<\/p>\n<\/div>\n

\n

Claude Mythos 5 and Fable 5 are built on the same foundation as Anthropic\u2019s Mythos Preview, which Anthropic dubbed too dangerous<\/a> to publicly release. (The company\u2019s warnings could be seen as genuine concern or more hype for their own model \u2014 or both.) Mythos 5 was made available to a select group of government agencies and companies, while Fable 5, which featured additional safeguards, was deemed \u201csafe for general use.\u201d But when a report indicated those guardrails may have failed, Anthropic\u2019s dire warnings about Mythos falling into the wrong hands came back to haunt it.<\/p>\n<\/div>\n

\n

A source familiar with the situation, who participated in the negotiations between Anthropic and the Trump administration, said the administration called the AI lab on Friday around 1pm ET and gave the company a 90-minute ultimatum to shut down access to Mythos 5 and Fable 5. If it didn\u2019t, then the government would impose export controls on Anthropic by authority of the US Commerce Department.<\/p>\n<\/div>\n

\n

The source said that Anthropic executives were talking to the White House within 15 minutes of that first call, confirming that CEO Dario Amodei joined the discussions about an hour and 15 minutes after that initial call. Amodei directly spoke with US Treasury Secretary Scott Bessent, Commerce Secretary Howard Lutnick, and National Cyber Director Sean Cairncross, in some cases more than once, the source confirmed.<\/p>\n<\/div>\n

\n

Anthropic wrote in a release on Friday that the company believed that the government \u201cbelieves it has become aware of a method of bypassing, or \u2018jailbreaking\u2019 Fable 5.\u201d Rather than an existential threat, though, Anthropic said that the jailbreak in question was a \u201cpotential narrow, non-universal\u201d one that was \u201cshared with the government\u201d by an entity the company declined to name. Moreover, Anthropic said the behavior wasn\u2019t unique to Fable 5. \u201cWe have reviewed a report that we believe is the basis of the government\u2019s directive and validated that the level of capability displayed there is widely available from other models (including OpenAI\u2019s GPT-5.5),\u201d Anthropic wrote.<\/p>\n<\/div>\n

\n

Semafor<\/em> reported<\/a>, citing one source familiar, that the hubbub began because the US government was concerned that a China-linked group had accessed the technology. But the source said that the China rumors went back weeks, referring to a large global telecommunications company that was initially cleared to be included in access to Mythos Preview, and that when the US government shared its concerns, Anthropic immediately revoked access.<\/p>\n<\/div>\n

\n

An X post<\/a> by David Sacks, the US government\u2019s former AI and crypto czar who stepped down in March, didn\u2019t mention China either. Sacks did, however, mention the unnamed entity that had exposed the issue to the government, calling it \u201ca highly credible trusted partner of both Anthropic and the USG who was testing Fable [which] came forward with a jailbreak of those guardrails.\u201d<\/p>\n<\/div>\n

\n

Some reports<\/a> point to Amazon CEO Andy Jassy as the person who flagged concerns to the US government after researchers at Amazon had red-teamed Fable 5. That conclusion stands at odds with some independent red-teamers, who have said<\/a> they were impressed with the level of the protections.<\/p>\n<\/div>\n

\n

The source familiar with the negotiations said that the Amazon research was explicitly mentioned in conversations with the US government. The person added that Anthropic had had access to that paper within days of the Friday export control directive and had been going back-and-forth since then with Amazon researchers to discuss it.<\/p>\n<\/div>\n

\n

Everything in that paper, the source said, could be achieved by OpenAI\u2019s GPT-5.5.<\/p>\n<\/div>\n

\n

Anthropic spent the weekend scrambling to make nice with the Trump administration, beginning with virtual meetings and then flying employees to DC, including Dave Orr, Anthropic\u2019s head of safeguards; Logan Graham, who runs its frontier red team and has led work on Project Glasswing; and Nicholas Carlini, a leading frontier developer and cybersecurity researcher. Axios<\/em> reported<\/a>, citing a source familiar with the Trump administration\u2019s thinking, that the company simply has repeatedly made missteps in its communication with the administration and that it \u201chas not done a great job at trying to speak to the administration and appreciate the ideological differences.\u201d For Anthropic, the timing couldn\u2019t be worse: the company had banked on Mythos to help it recover<\/a>, in part, from months of high-profile clashes with the US Department of Defense.<\/p>\n<\/div>\n

\n

The source familiar with the negotiations said that Anthropic pre-briefed the administration on Fable 5, and that the US Department of Commerce conducted testing pre-deployment, with no concerns shared at the time. The source added that Anthropic had been working closely with government agencies since Mythos Preview\u2019s release.<\/p>\n<\/div>\n

\n

The Trump administration initially took a hands-off approach to AI safety \u2014 but post-Mythos, it has become more ambivalent<\/a>, even as it frets over the threat of losing the AI race to China. Now, prominent cybersecurity leaders have warned<\/a> that sidelining Mythos 5 and Fable 5 could give China a significant AI advantage. Trump\u2019s move has galvanized international calls<\/a> for alternatives to American AI systems, while effectively putting a major US AI company\u2019s new flagship model on ice.<\/p>\n<\/div>\n

\n

A public letter<\/a> from tech and cybersecurity executives called for restrictions on Fable 5 to be repealed on Sunday. \u201cNot all of us agree that AI regulation is the right way forward,\u201d the letter states, adding that if regulations are going to happen regardless, then they should be rooted in \u201cscientific evaluations developed with input from industry and academia.\u201d<\/p>\n<\/div>\n

\n

Alex Stamos, chief product officer at Corridor, told The Verge<\/em> he organized the public letter because the countless number of vulnerabilities in the past decade-plus, written in a variety of different coding languages, require AI to patch before bad actors find them. \u201cWe\u2019re in a race, and I think policymakers don\u2019t understand that,\u201d Stamos said. \u201cThere\u2019s this weird arrogance, this idea that American labs are hugely ahead of our adversaries that will always be true, that it\u2019s really important to restrict access because of that. I just think that\u2019s foolish. If the labs are ahead, it\u2019s only by a matter of months. And you can see that in the open evaluations. The cutting-edge models are only something like six months ahead of the Chinese models \u2014 and those are the models we know about.\u201d<\/p>\n<\/div>\n

\n

The public letter goes on to state that though Anthropic\u2019s Mythos-class models are skilled at finding cybersecurity vulnerabilities and taking advantage of exploits, they aren\u2019t \u201cuniquely good\u201d at these tasks and that Fable 5\u2019s safeguards \u201cwere so aggressive as to be the source of humor in the cyber community on launch day.\u201d Stamos told The Verge<\/em> that \u201cthere\u2019s a real overstatement of Mythos\u2019 capabilities. Anthropic is somewhat responsible for this themselves, clearly \u2026 Mythos is great, but the real turning point was really last year.\u201d<\/p>\n<\/div>\n

\n

Stamos said the industry is awash with backup contracts being signed with non-US companies and open-weight models being deployed on alternative hardware arrangements because the past weekend made political risk part of companies\u2019 business plans more than ever before.<\/p>\n<\/div>\n

\n

\u201cThey are laughing at us in Beijing right now,\u201d Stamos said. \u201cOne of America\u2019s champions is being kneecapped by the US government while we\u2019re in a race with the Chinese. It\u2019s just incredibly stupid. That\u2019s why I wrote the letter, and I think that\u2019s why a lot of people signed onto it.\u201d<\/p>\n<\/div>\n

\n

Ben Van Roo, co-founder and CEO of Legion Intelligence, a system of agents for the national security community, told The Verge<\/em> that \u201cthe directive of \u2018no foreign national should use this model\u2019 is the most impossible thing to enforce.\u201d He added, \u201cWhen I first read that, my whole\u2026 [network of] AI community nerds was exploding.\u201d<\/p>\n<\/div>\n

\n

To make matters even more urgent, OpenAI, Google, and Microsoft have all come out with their own comparable products to Anthropic\u2019s Mythos, making many of the same claims about their effectiveness and risks. If the Trump administration bans Anthropic\u2019s advanced cybersecurity models, it can make a case for banning its competitors\u2019 models, too. That could spur AI industry leaders to unite and help out Anthropic or, as with its fight over autonomous weapons with the Pentagon, position themselves as a safer and more compliant alternative<\/a>.<\/p>\n<\/div>\n

\n

Even as the Trump administration is trying to free tech companies<\/a> of regulatory hassles, the Anthropic order could amount to a dramatic restriction on powerful AI models \u2014 depending on how the next few days play out.<\/p>\n<\/div>\n

\n

Legion Intelligence\u2019s Van Roo called it \u201cuncharted territory\u201d in the regulatory setting, adding that he doesn\u2019t think this is the last time something like this will happen.<\/p>\n<\/div>\n

\n

We\u2019ve also entered the era of AI populism, when a growing number of people are pushing back against the AI industry\u2019s outsized influence and the concentration of power at the top via data center protests, pledges to quit<\/a> using AI chatbots, lawsuits over wrongful deaths, and even attempted attacks on AI company CEOs. Van Roo says the Trump administration\u2019s recent moves against Anthropic could stoke \u201cgreater fears and concerns, potentially for the wrong reasons.\u201d<\/p>\n<\/div>\n

\n

The source familiar with the negotiations described the weekend\u2019s conversations as constructive, with some members of the administration admitting that putting export controls on model providers isn\u2019t ideal, since competitors with similar products may find themselves under the same restrictions \u2014 and since the US government is currently exploring<\/a> a program that would encourage the export of American AI systems.<\/p>\n<\/div>\n

\n

As Anthropic continues to negotiate with the US government, there\u2019s little chance that the company\u2019s other myriad issues with the Pentagon won\u2019t come up \u2014 namely, the ongoing battle between Anthropic and the Department of Defense over acceptable usage policies for Anthropic\u2019s tech by the US military.<\/p>\n<\/div>\n

\n

\u201cThis is new and we\u2019ve never had anything potentially this drastic before, and it does have some real ramifications\u201d in terms of how to enforce access to powerful models, Van Roo said. \u201cWho gets to use this new technology that continues to outpace our own ability to regulate it?\u201d<\/p>\n<\/div>\n

Follow topics and authors<\/strong> from this story to see more like this in your personalized homepage feed and to receive email updates.<\/span><\/p>\n