On April 11, according to CNBC, before the AI company Anthropic released its latest powerful model "Mythos", US Vice President Vance and Treasury Secretary Bessent held a meeting with the CEOs of a number of leading technology companies last week.On issues such as the security of AI models and how to deal with cyber attacksThey were questioned.
bessant and vance
The meeting was held by phone, according to two people familiar with the matter. Technology leaders participating in the meeting included Anthropic CEO Dario Amodei, xAI CEO Elon Musk, Google CEO Sundar Pichai, OpenAI CEO Sam Altman, Microsoft CEO Satya Nadella, CrowdStrike CEO George Kurtz and Palo Alto Networks CEO Nikesh Arora.
People familiar with the matter said the purpose of the meeting between CEOs of these technology companies was toDiscuss the security posture and security deployment issues of large language models. Officials at the meeting also discussed how to respond if the development of the model evolves in a direction that is beneficial to attackers, such as cyber attacks.
An Anthropic official declined to comment on the meeting, but told CNBC on Friday that the company has been in communication with senior government officials about cybersecurity issues in recent weeks and has expressed a willingness to support "the government's testing and evaluation of this technology."
"Before releasing the model to the public, Anthropic has briefed senior officials from various U.S. government departments on the full capabilities of the Mythos preview version, including its application in cyber attack and defense. From the beginning, our priority has been to let the government understand the situation as early as possible, including what the model can do, where the risks are, and how we manage these risks." Anthropic official said.
OpenAI and CrowdStrike declined to comment on the meeting. The White House and other related companies have not yet responded to CNBC's request for comment.
Anthropic launched its Mythos AI model to limited partners on Tuesday while it evaluates measures to prevent hackers from exploiting the model. Apple, Google, Microsoft, Nvidia, Palo Alto Networks and CrowdStrike are among its first launch partners.
Bessant and Federal Reserve Chairman Jerome Powell unexpectedly convened a meeting this week with the heads of major U.S. banks to discuss potential threats posed by the Mythos model, a sign of the Trump administration's deepening concerns about advanced cyber tools.
It also shows that Anthropic remains involved in White House conversations about AI even as President Trump seeks to remove Anthropic's Claude platform from federal agencies. Anthropic's legal challenge to the U.S. Department of Defense's determination of its supply chain risks continues in court, following two competing rulings from courts on both coasts of the United States.