Scott Bessent and Jerome Powell held a closed-door meeting with top U.S. bank executives to raise concerns about potential cybersecurity threats tied to a newly developed artificial intelligence model by Anthropic, according to sources familiar with the matter.
Truth Live News International gathered that the discussions focused on the company’s latest system, known as Mythos, which has demonstrated the ability to identify and exploit vulnerabilities across major operating systems and web browsers. Officials warned that such capabilities could pose significant risks to financial institutions if not properly managed.
The meeting, hosted in Washington, was aimed at ensuring that major banks are aware of emerging AI-related threats and are strengthening their cyber defenses accordingly. Invitations were extended while several banking executives were already in the capital for scheduled engagements.
Access to the model is expected to remain restricted, with only a limited number of technology firms, including Microsoft and Google, being considered for early use.
Sources indicated that Anthropic had also engaged with U.S. government officials prior to the meeting, outlining both the defensive and potentially offensive cyber capabilities of the system.
Several financial institutions and regulatory bodies declined to comment on the discussions, while no official statement has been issued regarding any immediate policy actions.
