AI agents may soon help people manage their finances, but major US banks are putting limits on how these tools access customer accounts. JPMorgan Chase and Charles Schwab are restricting unauthorized AI agents that attempt to access customer accounts using banking credentials, citing security and financial risks.
The move comes as technology companies develop AI assistants capable of handling tasks on behalf of users, including managing money and finding better returns on savings. However, banks are concerned about giving outside software access to sensitive financial information and the ability to initiate transactions. The restrictions could complicate the expansion of autonomous financial assistants.
Why are Banks Blocking AI Agents?
JPMorgan Chase CEO Jamie Dimon has warned that allowing unapproved AI agents to access customer accounts could expose both customers and the bank to unnecessary risks. The bank is particularly concerned about third-party tools that use customers' usernames and passwords to log in and operate accounts. JPMorgan CFO Jeremy Barnum has also criticised this approach, describing credential sharing and screen scraping as insecure ways to access financial information.
Charles Schwab has taken a similar position. CEO Rick Wurster said the brokerage blocks unauthorized agents attempting to log in on behalf of clients. The company maintains that these restrictions are intended to protect customers and its own systems.
Could AI Agents Move Money Away from Banks?
Security is not the only concern. Banks and investors are also watching the possibility of AI assistants automatically moving money from low-interest accounts into alternatives offering better returns.
For example, an AI agent could compare savings accounts and transfer idle cash to another financial institution offering a higher interest rate. If large numbers of customers adopt such tools, banks could lose deposits that help fund their lending activities.
However, this scenario remains largely a concern about future adoption rather than an established trend. Wells Fargo CEO Charlie Scharf has questioned whether consumers will be willing to give autonomous agents broad authority over their money.
Banks are Developing Their Own AI Tools
While restricting outside agents, financial institutions are continuing to explore controlled uses of artificial intelligence. Schwab has introduced Charley, an AI assistant designed to answer customer questions and perform limited tasks, including changing account details. The brokerage is also testing AI tools for financial advisers and selected traders.
JPMorgan is exploring ways to provide AI-powered services while maintaining control over account access and customer data. These efforts suggest that banks are not rejecting AI altogether; they want greater oversight of how the technology interacts with financial systems.
Also Read: AI Agents, Zero-Click Attacks: The Next Cybersecurity Threat
What this Means for the Future of AI Banking
The restrictions point out a potential conflict between the convenience offered by artificial intelligence and financial security. Self-sovereign agents could ease personal finance management for people, but problems like errors, unauthorized actions, and uncertainty about accountability persist.
Customers should not assume that providing login credentials gives an AI assistant unrestricted access to their bank accounts. Future implementations may rely on secure, authorized access methods that allow banks to control and monitor an AI agent's capabilities.