KEY POINTS
- Wells Fargo's updated customer agreement, effective Tuesday, says use of any AI agent is at the customer's own risk.
- JPMorgan Chase banned third-party agents that log in with customer credentials, and Bank of America, Citibank and PNC issued warnings or changed terms.
- American Express said some agent errors may become eligible for reimbursement if a customer cannot recover money from the retailer.
Wells Fargo is rolling out an updated agreement for millions of customers that focuses on AI agents, starting on Tuesday, WIRED reported. The new terms state that using any AI agent is done "AT YOUR OWN RISK," placing responsibility for automated mistakes on the customer.
Wells Fargo is not alone. According to WIRED, JPMorgan Chase changed its terms in late October to ban third-party agents that sign in with a customer's credentials. On the same day, Bank of America warned customers not to share login details with agents. Citibank and PNC have also updated their terms to include warnings about the tools.
AI agents, which automate online tasks and click on a user's behalf, have had a bumpy 2026. A growing group of power users has adopted them, with mixed results. Some have benefited from the automation, while others have been embarrassed by serious errors, such as an agent posting bank details where coworkers could see them.
The pushback extends beyond banking. WIRED said Amazon has blocked Meta's Muse, one of the most widely used agents, from making purchases on its platform. United recently announced that personal bots would not be allowed to buy flights for users.
Experts quoted by WIRED see the bank moves as precautionary. Sucharita Kodali, a Forrester analyst, said most people do not even know what agents are, and she described banks as fundamentally risk-averse. She pointed to concerns ranging from transaction security to automated withdrawals that could harm account holders. Kacy Kim of Bryant University said consumers authorize goals, not every click, and that some may find it unfair to be held accountable for an agent's errors. She does not expect the policies to deter committed early adopters.
Luca Cian, a University of Virginia professor, said shoppers are comfortable letting AI handle dull purchases such as household supplies, but are warier of big-ticket buys. He offered a hypothetical in which an agent accidentally orders 10 roasted chickens, noting that the retailer, the AI developer and now the bank can each argue they are not responsible.
Some card providers are exploring limited protection. American Express announced that certain agent mistakes, such as ordering too much of an item, may qualify for reimbursement in the future if the shopper cannot get their money back from the retailer.
WIRED noted that banks are also building their own AI tools. Cian said banks are locking the door to outside agents while developing their own. Wells Fargo's generative AI assistant has been used more than a billion times, Bank of America revamped its EricaAssist agent for employees this year, and Citigroup launched an internal platform centered on agents. Kodali questioned whether agentic shopping is needed at all, since buying from major retailers is already easy.