[Anthropic's Mythos Model Detected Vulnerabilities in U.S. Government Classified Systems During Testing] According to Foresight News and a report by the Associated Press, a U.S. official stated on Tuesday that Anthropic's Mythos model identified vulnerabilities in highly sensitive and secure computer systems of the U.S. government during a test. The official noted that the test was conducted as part of the 'Glass Wings Project,' initiated by the human factors company. This project brings together tech giants and other companies with the aim of protecting critical global software from the potentially 'severe' impacts of the Mythos model on public safety, national security, and the economy.