Threats to US payment rails helped trigger Bessent's AI worries (www.semafor.com)

Treasury Secretary Scott Bessent has taken a proactive stance on artificial intelligence policy due to increasing concerns from financial institutions about the impact of advanced AI models like Anthropic's Mythos on critical infrastructure, particularly the U.S. payment rails. Banks have long warned that vulnerabilities in these systems, which facilitate rapid and extensive monetary transfers, could lead to significant disruptions in the financial system. In recent meetings with the Federal Reserve and banking leaders, Bessent emphasized the need for a balance between innovation and safety to maintain the U.S.'s competitive edge over countries like China. Bessent's involvement has led to a shift from a previously hands-off approach to a more regulated framework, including an executive order that outlines a voluntary review process for advanced AI models. This initiative aims to ensure greater oversight of AI technologies that could impact financial institutions and infrastructure. However, negotiations with the Biden administration regarding the AI models' operational status, due to export controls imposed by the Commerce Department, might delay these developments. The outcome of these discussions will be crucial, given that the Fed wired over $283 trillion through its payment systems in just the first quarter of 2026, highlighting the potential systemic risks associated with AI vulnerabilities.