Vance, Bessent questioned tech giants on AI security before Anthropic's Mythos release

In a recent teleconference, Vice President JD Vance and Treasury Secretary Scott Bessent engaged with influential tech executives about the security measures surrounding artificial intelligence models, particularly in light of Anthropic's impending release of its Mythos AI model. This private meeting, attended by key figures such as Anthropic's Dario Amodei, Google's Sundar Pichai, OpenAI's Sam Altman, Microsoft's Satya Nadella, CrowdStrike's George Kurtz, and Palo Alto Networks' Nikesh Arora, focused on the safety and deployment strategies for large language models. The discussion also covered potential responses to cyber threats should these AI models be exploited by malicious actors. While OpenAI chose not to comment on the meeting, attempts to gather insights from the White House and other participating companies were also made. An Anthropic representative confirmed that the company has been liaising with government officials regarding cybersecurity matters and has offered support for the government's evaluation of the new technology. The representative stated, "Before any external rollout, Anthropic briefed senior officials across the U.S. government on the full capabilities of Mythos Preview, including its offensive and defensive cyber applications." They emphasized the importance of involving government stakeholders early in the process to discuss potential risks. On Tuesday, Anthropic officially launched the Mythos AI model to a select group of companies, including tech giants such as Apple, Microsoft, and Nvidia, as part of its strategy to mitigate the risk of misuse by hackers. In a related move, Bessent and Federal Reserve Chair Jerome Powell convened an unexpected meeting with leaders from major U.S. banks to discuss the implications of Mythos, reflecting growing concerns from the Trump administration regarding advanced cyber capabilities. Anthropic remains a focal point in the ongoing AI discourse within the government, particularly as President Trump aims to remove the startup’s Claude platform from federal use. The company is currently embroiled in legal battles regarding the Department of Defense's supply chain risk designation, with mixed rulings complicating its ability to secure DOD contracts, though it continues to collaborate with other federal entities while the legal matters unfold.