Policy
Treasury Nominee Declines to Say Who's Liable for Rogue AI Agents
Rebekah Jurata told senators there are 'unanswered questions' as banks shift responsibility to customers for third-party AI account access.
Key takeaways
- Treasury nominee Rebekah Jurata declined to specify liability rules for unauthorized AI agent transactions, citing "unanswered questions" as the technology develops.
- The Bank Policy Institute said Friday that AI agents' defenses against prompt injection attacks are "not nearly good enough" to justify account access.
- Sen. Ruben Gallego warned that AI agents searching for risk could trigger bank runs that "could really cripple our economy."
- Treasury Secretary Scott Bessent previously told Congress that AI creators should be "liable for what they build and generate," contrasting with Jurata's noncommittal stance.
- Financial institutions are increasingly pushing back against third-party AI account access, citing insufficient risk management protocols.
A senior Treasury Department nominee declined to specify who should bear responsibility when artificial intelligence agents make unauthorized transactions in consumer bank accounts, telling senators the issue requires further study as the technology evolves.
Rebekah Jurata, nominated for assistant secretary at Treasury, appeared before the Senate Banking, Housing and Urban Affairs Committee on Thursday where she faced pointed questions about accountability for AI-driven financial activity. When Sen. Ruben Gallego (D-Ariz.) asked who should be held responsible if a "rogue AI agent" makes an unauthorized transfer, Jurata called the question "incredibly timely" but stopped short of taking a definitive position.
"What you want to balance is the ability of everyday consumers to take advantage of these new products and tools to see if they can have enhanced return on their investment and whatnot," said Jurata, who previously served as general counsel at the American Investment Council and as deputy assistant secretary at Treasury during the first Trump administration. "But also we want to maintain the resiliency and the safety and soundness of the banking system."
Why it matters
The hearing comes as financial institutions are beginning to draw hard lines against third-party AI access to customer accounts, citing insufficient risk management protocols. The Bank Policy Institute said in a Friday blog post that agents' large language models aren't adequately protected against prompt injection attacks that could allow fraudsters to drain accounts. With AI agents increasingly capable of autonomous financial decisions, the question of liability—whether it falls on consumers, banks, AI developers, or some combination—has immediate implications for both financial stability and consumer protection.
Systemic risk concerns
Gallego pressed further, asking what should happen if AI agents designed to identify risk trigger bank runs—a scenario he warned "could really cripple our economy." Jurata responded that Treasury is "more optimistic as it relates to opportunities that super intelligence is providing to the banking system," pointing specifically to cybersecurity improvements. The Trump administration uses "super intelligence" as its preferred term for AI.
When Gallego asked a third time about specific guardrails Treasury believes should govern AI in finance, Jurata committed to "coordination" with other bank regulators within the "existing legal framework" to "both embrace the use of this new technology while also maintaining the integrity of the system."
Contrasting positions within Treasury
Jurata's cautious stance contrasts with comments Treasury Secretary Scott Bessent made to the House Financial Services Committee in September. Bessent came out strongly against liability exemptions for AI labs, telling lawmakers that the "best way to guarantee safety" is for AI creators to be held "liable for what they build and generate."
During Thursday's hearing, Jurata also addressed financial scams more broadly, telling Sen. Bill Hagerty (R-Tenn.) that Treasury is playing a "significant coordinating role" with federal regulators, law enforcement and the private sector on the issue. She told Sen. Bernie Moreno (R-Ohio) that "a holistic solution" is needed to address fraudulent advertisements on social media.
The details were first reported by FedScoop.