The White House advances the opening of Mythos to federal agencies On April 17th, according to a memo, the US government is preparing to provide a version of Anthropic's artificial intelligence model Mythos to major federal agencies, while strengthening security measures, to address concerns that the tool may significantly increase cybersecurity risks. In an email sent to various cabinet departments on Tuesday, the Federal Chief Information Officer of the White House Office of Management and Budget (OMB), Barbasia, stated that OMB is establishing relevant protective measures to enable agencies to start using this highly restricted AI tool in the future. The email did not specify whether each institution could access Mythos, nor did it provide a specific timetable or usage method. The email only informs the heads of various technology and cybersecurity departments that more information is expected to be received in the coming weeks. At present, Anthropic only opens Mythos to a few technology companies and financial institutions, and encourages them to use it to assess their own cybersecurity risks. Due to concerns that hackers may use the model's capabilities to steal data or damage the system, the company has imposed strict restrictions on its release scope.