Banking and trading sectors face AI agent risk and liability questions

Traders Union
All News
Financial Institutions
Banking, trading AI risks

Highlights

FDIC chairman Travis Hill flagged open questions about how much decision-making authority consumers will delegate to AI agents in banking and trading.
Alpaca CEO Yoshi Yokokawa identified liability risks if AI agents act improperly or cause losses, prompting his fintech to limit fully automated trading.
Current FDIC model risk management rules do not cover agentic or generative AI, so banking sector adoption depends on future regulatory guidance and developing trust frameworks.

Regulatory and industry concerns emerge

As reported by Axios, speakers at the Las Vegas event said the spread of AI agents in finance is opening a new set of operational and legal challenges as the technology gains more authority over money movement and trading decisions.

Federal Deposit Insurance Corporation chairman Travis Hill said advanced technology is helping make banking more efficient, noting that consumers can already move money easily between banks. He said a key open question is how far depositors are willing to delegate decision-making authority to AI agents as those tools become more capable.

Alpaca co-founder and CEO Yoshi Yokokawa said liability remains a central issue if agents act improperly, engage in criminal activity or execute bad trades that lose money. He also said fully relying on agents to generate profits for consumers is dangerous, underscoring the limits that his fintech platform places on fully automated trading.

Rules and trust remain in development

Hill said the debate over AI regulation is still evolving and explained that the FDIC's model risk management guidance does not apply to agentic AI and generative AI as the agency considers future guidance. He said regulators are trying to stay current with rapid technological change while also taking a deliberate approach.

The discussion highlights a broader challenge for banking and trading firms as they test AI-driven services, because faster and more automated customer experiences may be easier to build than the trust framework needed to support them. For the financial sector, that means adoption is likely to depend not only on product capability but also on clearer rules around responsibility, consumer protection and risk control.

In our earlier coverage of the Bank of England’s warning on AI-driven cyber risk, we explained why regulators are urging banks to plan for severe tech failures and build deeper resilience for critical services. The article highlighted “bare metal” recovery options that can keep core banking functions running if primary systems are compromised, reflecting concern that AI is increasing the scale and complexity of attacks.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.