Vice President JD Vance and Treasury Secretary Scott Bessent convened a private conference call with leading AI executives in early April to probe concerns over artificial intelligence model security and cyber attack responses.

The discussion occurred just a week before Anthropic released its advanced Mythos model, which the company limited to about 40 major tech firms due to potential hidden cybersecurity risks. Participants included Anthropic CEO Dario Amodei, Alphabet CEO Sundar Pichai, OpenAI CEO Sam Altman, Microsoft CEO Satya Nadella, and leaders from cybersecurity firms Palo Alto Networks and CrowdStrike.

Details of the call emerged from sources familiar with the matter, who spoke on condition of anonymity because the event was private. The Trump administration officials focused on how AI models like Mythos could expose vulnerabilities or enable sophisticated cyber threats, prompting questions about preparedness measures.

Anthropic had been in ongoing talks with senior U.S. government figures ahead of the launch, highlighting the model's dual-use potential in both bolstering defenses and aiding attacks. Separately, Bessent and Federal Reserve Chair Jerome Powell met with heads of major U.S. banks to address similar risks posed by Mythos.

The closed-door nature of the session underscores the administration's hands-on approach to balancing AI innovation with national security, without resorting to heavy regulations that Vance has previously criticized. Companies involved, including Anthropic, Alphabet, OpenAI, Microsoft, Palo Alto Networks, and CrowdStrike, declined to comment when reached by reporters.

Mythos represents a leap in AI capabilities, but its restricted rollout reflects industry caution amid escalating concerns over AI's role in cybersecurity. The model can identify vulnerabilities faster than human experts, raising fears it could be weaponized by adversaries.

This meeting aligns with broader White House efforts to position the U.S. as the global leader in safe AI development. Earlier this year, Vance addressed the Paris AI Action Summit, advocating light-touch policies to avoid stifling growth. The private briefing signals continued vigilance as AI advances accelerate.