Secretary Scott Bessent told Congress this week that the Treasury Department is focusing on public-private partnerships and potentially an AI sandbox to drive the adoption of artificial intelligence across the financial services sector in a “gradual” but ultimately “powerful” manner.
Bessent appeared before the House Financial Services Committee on Wednesday and the Senate Banking Committee on Thursday, where he was pressed by several lawmakers to address the AI priorities set out in the Financial Stability Oversight Council’s annual report to Congress.
A report released by the Treasury Department in December listed “Using Artificial Intelligence to Promote Financial Stability” as one of FSOC’s four key focus areas. The report recommended that agencies use the FSOC’s AI working group to identify “regulatory hurdles” to financial institutions’ adoption of AI.
Sen. Mike Rounds (R.S.D.), co-chair of the Senate AI Caucus, asked Bessent what he believes are the “biggest obstacles” preventing banks from “responsibly implementing AI, especially in compliance, fraud detection, and risk management.”
Bessent said there is still “a lot of work to do” when it comes to AI, and the approach so far has been for regulators to “work with private partners to implement it in a phased manner”, which will lead to “robust use of AI”.
He added that while “AI can be a great tool,” “we also have to consider that AI can be a risk through state and non-state actors. So this is a public-private partnership that we are pushing very hard across the agency and across the Department of Treasury.”
Last July, Mr. Rounds introduced a bill that would direct the Securities and Exchange Commission, Federal Reserve Board, Consumer Financial Protection Bureau, and other federal financial institutions to establish internal AI innovation labs. According to the bill’s language, these labs would essentially serve as sandboxes for government agencies to test AI projects “without the expectation of unnecessary or unduly burdensome regulatory or enforcement action.”
Rounds asked Bessent whether a “time-limited AI sandbox” for financial institutions could help them safely test AI tools while giving regulators a chance to assess risk.
“That’s obviously one of the very interesting options and we’re looking at that going forward,” Bessent replied. “I would be happy to work with your staff.”
The Treasury Secretary similarly told House Financial Services Committee Chairman French Hill (R-Ark.) that he is ready to work with lawmakers to “ensure that technology does not outpace the bill.” Hill specifically asked for the FSOC’s views on how AI can deliver customer service in financial services, along with “more robust compliance processes” across the banking and securities industries.
“There are two elements to this,” Bessent told The Hill. “For example, there are service improvements that we’re working on at the IRS, where we want to leverage artificial intelligence to help ease the burden on customers and reduce service. [there is] Ensure financial security and alert everyone to the risks of what is happening. But this is a transformative change. ”
Part of that change has to do with how the Treasury Department views AI as a means to strengthen its cybersecurity posture. In an exchange with Rep. Josh Gottheimer (DN.J.) and Rep. Andrew Garbarino (RN.Y.), Bessent said Treasury is working closely with its partners in the financial sector and inviting them for tabletop exercises and other “meetings” to ensure everyone is working from the same strategy.
“I think it’s important to work together,” Bessent told Gottheimer. “A lot of times in these very fast technology cycles we see technology preempting regulation, so whether it’s in the financial system or anywhere else, it’s important that we work together to keep regulation and technology in sync. … So, again, it’s about building resilience and being aware of best practices and working with our private sector partners.”
