According to reports, executives of AI startup Anthropic held their first in-person meeting with senior officials of the Trump administration on Monday. This followed an export ban imposed by the federal government late Friday evening citing security vulnerabilities, which prohibited non-US users from accessing its latest model, Fable 5, forcing the company to fully delist it. The ban originated after major investor Amazon discovered methods to bypass Fable’s safety guardrails and reported them to the White House and the National Security Agency. Over the weekend, Anthropic co-founder Tom Brown had multiple calls with Commerce Secretary Howard Lutnick and National Cyber Director Sean KaneCross. At Tuesday’s technical briefing, Anthropic’s security team detailed its cybersecurity defense mechanisms. This marks the first time the White House has forced a company to withdraw an AI model. White House officials stated that lifting the ban might take more than several days but noted “it depends on Anthropic,” leaving the window open. Treasury Secretary Bessent led dialogues prior to the ban but reduced his involvement over the weekend; both he and Lutnick have since traveled to France with the President for the G7 summit. In a blog post, Anthropic argued that the vulnerability was “relatively simple” and could be found in other publicly available models. Nearly 80 tech leaders signed a joint letter to the White House over the weekend, arguing that such capabilities were “essential skills for writing secure code” and that export controls would “remove the best models from defenders’ hands.” Some government insiders also warned that if the dispute dragged on, it could de facto create an approval regime for AI models, hindering innovation and China competition dynamics.