Anthropic says US government forced it to disable cybersecurity AI models
Anthropic says US government forced it to disable cybersecurity AI models
https://therecord.media/anthropic-says-gov-forced-it-to-disable-cyber-ai-models
Publish Date: 2026-06-15 08:38:00
Source Domain: therecord.media
Using an unordered list, summarize the following article with between 4 and 8 key points. Anthropic disabled two of its most advanced, cybersecurity-focused AI models on Friday, saying it was responding to an export control directive issued by the U.S. government barring any foreign national from accessing them. The directive covered foreign nationals inside the United States as well as abroad, including Anthropic’s own employees. In a statement, Anthropic said the “net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.” The directive itself has not been made public. According to the company, it cited national security authorities. It appears to be the first time such authorities have been used to curtail the export of AI models rather than chips or hardware. Anthropic said its understanding is “that the government believes it has become aware of a method of bypassing, or ‘jailbreaking’ Fable 5,” although it said the government provided only verbal evidence of this technique. The company reviewed what it believes is the underlying report and found the vulnerabilities identified were minor, previously known and reproducible using other publicly available models including OpenAI’s GPT-5.5. Anthropic said it was complying with the directive while disputing its basis: “We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.” “If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers,” the company wrote. The directive was issued two days after Anthropic’s chief executive Dario Amodei published a policy essay calling on the government to have legal authority to block unsafe AI deployments. On Friday, the company said while it supports giving the government authority to block unsafe AI deployments, this should be done through “a statutory process that is transparent, fair, clear, and grounded in technical facts. This action does not adhere to those principles.” The directive arrives against a backdrop of escalating tensions between Anthropic and the Trump administration. In February, Defense Secretary Pete Hegseth designated Anthropic a “supply chain risk” — a label historically applied to companies such as Huawei — after contract negotiations over military use of Claude broke down. Earlier this month, Anthropic filed for an initial public offering with investors anticipating it will value the company above $1 trillion, as reported by the Financial Times. The company said the draft registration statement gave it the option to go public, depending on market conditions and other factors. Anthropic apologized for the disruption to customers and said “We believe this is a misunderstanding and are working to restore access as soon as possible.” Get more insights with the Recorded FutureIntelligence Cloud.Learn more.
