US Commerce Secretary Howard Lutnick revealed that the government's action against Anthropic's Fable 5 and Mythos models was driven by fears they could be deployed by military intelligence users in China, Russia, or other countries of concern. According to a letter sent to Anthropic CEO Dario Amodei on Friday, Lutnick ordered the company to suspend export of the AI models to destinations worldwide and all foreign nationals, wherever located. Senior Anthropic technical staff met with officials at the Department of Commerce in Washington on Monday to negotiate a solution, with Commerce Secretary Lutnick holding regular calls with Anthropic officials as he works toward a deal. National Cyber Director Sean Cairncross joined Monday's working-level meeting with Anthropic at the Commerce Department, as reported by Reuters. The stakes remain high, with the government seeking assurances that the models cannot be used to harm the US, while Anthropic is pushing to restore access to its top-tier models after taking them offline for all users on Friday.
Anthropic has suspended access to Fable 5 and Mythos 5 AI models following a U.S. export control directive received at 5:21 PM ET on Friday evening. According to CNBC reports, the company received a directive from Secretary Howard Lutnick under the Bureau of Industry and Security requiring it to block all foreign nationals from accessing its most powerful AI models. The order, issued as a mandatory licensing action, requires Anthropic to file additional applications for individually validated licenses and carries financial and civil penalties for non-compliance. In a blog post, Anthropic revealed that the U.S. government told the company it believes there is a method of bypassing, or "jailbreaking," a safeguard against using the model to find cybersecurity holes. The bypass found only "minor" security flaws that other publicly available models can also find. The letter to Anthropic said the Commerce Department was taking action through authorities granted it under the 2018 Export Control Reform Act to impose controls on emerging technologies essential to US national security, marking the first time the Commerce Department has used that power according to export control experts.
The government's swift action came after Anthropic researchers reported that Fable 5 had been pushed to supply restricted cyberattack information. According to the Wall Street Journal, Amazon is one of Anthropic's largest backers, having invested billions of dollars. The firm's report showed Fable 5 surfaced security bugs in at least four software programs when fed a specific set of queries. Jassy raised these security concerns with senior Trump administration officials, with White House officials soon convening to map out a response. The Pentagon's chief information officer appeared to endorse the move, writing on X that "some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation."
The government's swift action came after Anthropic CEO Dario Amodei published an essay this month warning that artificial intelligence's power "has become undeniable" and presented "very real risks" to cybersecurity, the financial sector, critical infrastructure, and national security. As reported by Live Mint, Amodei positioned these dangers as "present and pressing" rather than distant theoretical concerns, arguing that concerns by lawmakers were out of step with AI's rapid progress. Earlier this month, Anthropic went further still, calling for a temporary halt to the development of frontier AI models in a research paper, warning that the latest generation of models was approaching the ability to improve themselves, which might increase the risks of humans losing control over AI systems. The company has publicly stated that the government should be able to block unsafe deployments through a process that is transparent, fair, clear and grounded in technical facts, arguing this directive "does not adhere to those principles."
More than 80 cybersecurity executives and experts signed an open letter to Lutnick and National Cyber Director Sean Cairncross on Sunday supporting Anthropic's position, with cybersecurity leaders at major firms including Nvidia and Adobe asking the Trump administration to lift the restrictions. Anthropic's technical staff have met with officials virtually every day since the Trump administration contacted the company on Friday, according to a person close to the company. Anthropic CEO Dario Amodei and Commerce Secretary Howard Lutnick are both set to attend the G7 meetings in Evian-les-Bains, France, where they may speak as negotiations continue. The San Francisco-based AI startup, which has confidentially filed for a US initial public offering, had previously warned about the hacking capabilities of its Mythos model and held it back from wide release, with the company working with the government to test Fable 5 before its release and receiving approval to deploy it.
The dispute has sent shockwaves through the AI industry, with Joelle Pineau, Chief AI Officer at Cohere Inc., warning that "the news from the last few days is a huge wake-up call for others across the industry." As reported by Bloomberg Television, Pineau noted that "the fact that models which a few days before were presented as some of the most capable in the industry suddenly are no longer accessible means that there can be a lot of volatility in terms of access to the technology." Ayham Boucher, executive director of AI Strategy and Innovation at Cornell University, emphasized that "describing this as a request to 'fix the guardrails' makes the problem sound much simpler than it is." Boucher explained that "jailbreak resistance is not like patching a normal software bug," highlighting the technical complexity of resolving security vulnerabilities in advanced AI models. The dispute marks the latest clash between the administration and the company, with Anthropic having been embroiled for months in a feud with the Pentagon over extra guardrails for military use, after contract talks broke down and the Defense Department declared the firm a supply-chain risk in March.
