Anthropic Pulls Its Most Powerful AI Models After U.S. Bars Foreign Access

Anthropic disabled its two most powerful AI models on Friday to comply with a U.S. government export-control directive prohibiting use by any foreign national. The directive blocked access to Fable 5 and Mythos 5 for all foreign nationals, including Anthropic employees, the company said. Users reported outages beginning Saturday as the company abruptly disabled access to the models. Anthropic framed the action as a necessary compliance measure, emphasizing that the controls aim to prevent foreign adversaries from leveraging its technology for national security risks. The move signals a broader push to treat AI models as strategic assets.

Officials described the action as part of a broader effort to curb foreign use of U.S.-made AI technology, similar to export controls once used for chips. The company has previously clashed with the administration over military applications, including a period when the Pentagon blacklisted Anthropic as too dangerous for government use. The Wall Street Journal cited a letter from Commerce Secretary Howard Lutnick to Anthropic's CEO, Dario Amodei, detailing the restrictions. Analysts cautioned that the timing could complicate an anticipated U.S. initial public offering, while rivals SpaceX and OpenAI discuss similar maneuvers.