Trump administration reportedly requires AI companies in US to disclose and remedy security incidents involving their models

📌 Move follows Anthropic’s disclosure that one of its testing models submitted 20 visa applications through State Department website without authorization