US banking authorities are turning their attention to artificial intelligence. AI can boost markets, but there are hazards that need to be closely watched, Treasury Secretary Scott Bessent told Congress.
Bessent was providing testimony on the Financial Stability Oversight Council’s 2025 annual report before the House Financial Services Committee.
AI is currently one of the top four priority for safeguarding the financial system, according to him. Cybersecurity, Treasury markets, and regulatory modernization are the other three.
Bessent stated that the council is giving the responsible application of AI top priority in order to improve financial stability.
Regulators are collaborating with both public and private partners, he said. Monitoring hazards and enhancing system resilience are the objectives.
According to Bessent, AI is rapidly gaining traction in financial markets and banks. Businesses use it for operations, compliance, and risk management. However, he cautioned that improperly comprehended or unmonitored use may exacerbate shocks under stressful situations.
According to him, the FSOC report shows a change in strategy. Broad warnings are becoming less common among regulators. Instead, they are concentrating on particular weaknesses that have the potential to cause systemic disruption.
“In order to focus on the issues that are most important for the financial stability of the United States, we are ignoring the noise,” Bessent stated.
Democratic members expressed concern over decisions made by AI. They cautioned that discrimination and bias could be introduced by opaque algorithms, particularly in credit and lending decisions. They claimed that previous financial crises had been exacerbated by similar blind spots.
Bessent recognized the dangers. The FSOC’s mandate, he argued, is limited. When asked about explainability in AI systems, he responded, “It could be a concern.” “However, we don’t think it’s important for financial stability.”
The administration’s strategy received broad support from Republican lawmakers. They argued that early regulation shouldn’t stifle innovation. They contended that AI can tighten compliance, enhance customer service, and aid in the detection of fraud.
Regulators are also keeping an eye on concentration hazards, according to Bessent. He claimed that an excessive dependence on a small number of AI suppliers or systems could increase the harm caused by technical malfunctions or hacks.
“Criminal organizations and nation-state actors continue to target our infrastructure and financial institutions,” he stated.
FSOC is coordinating with foreign peers, Bessent continued. He stated that in order to prevent regulatory loopholes, international cooperation is necessary to mitigate the global risks associated with AI.