Ally, Chase and Citi Tell Customers: You're Liable for Your AI Agent's Mistakes
Ally, Chase, Citi and Wells Fargo are warning customers they may be liable when AI agents make purchases or move money on their behalf, as autonomous shopping tools spread.
By Grace Kim
3 min read

What's News
- Ally Bank emailed customers this month that they are "responsible for any actions that the AI takes."
- Chase's November email warned that "authorized transactions may not be reversible if your information is misused."
- Amazon blocked Meta's Muse agents from purchases in September, citing safety concerns.
- JPMorgan CEO Jamie Dimon warned about AI agent risk on an October earnings call.
- Axios reported in October that major banks detected third-party agents inside customer accounts.
Ally Bank emailed customers this month with a blunt message: "When you use an AI service, you're responsible for any actions that the AI takes." Chase, Citi and Wells Fargo have sent the same warning in different words. The banking industry has decided that your AI agent is you — and its mistakes are your losses.
The alerts, reviewed by Business Insider across emails and app notifications, mark a shift in how banks treat a new class of AI tools that don't just answer questions but act. Agents can book flights, reserve restaurant tables, or buy deodorant before you run out. When they buy the wrong thing or send money to the wrong person, the account holder may have no recourse.
What exactly did the banks say?
Ally's email didn't mince words: "AI can make mistakes, hallucinate, misunderstand your instructions, or follow them in ways you didn't intend, which could lead to significant loss."
Chase warned customers in a November email that "authorized transactions may not be reversible if your information is misused." That single sentence carries the legal weight of the whole campaign — once a transaction counts as authorized, standard fraud protections may not apply.
Citi added a banner to its website and mobile app pointing users to a caution page: "AI assistants can make mistakes, and that means you can be liable for errors it makes."
The core message: an AI agent with access to your financial information will be treated as an extension of you, not as an outside intruder.
Why are banks sending these warnings now?
The timing tracks the arrival of autonomous agents. Earlier AI models like OpenAI's ChatGPT or Anthropic's Claude responded with blocks of text. Newer online agents — Meta's Muse, Instact AI, OpenAI's Dots — go out and complete tasks on a customer's behalf.
Retailers have already been forced to respond. In March, Business Insider reported that Target changed its customer agreement to state that AI-agent shopping was "considered transactions authorized by you." In September, Amazon blocked Muse agents from making purchases on its site, citing safety and reliability concerns.
Banks are also drawing a line between authorized account connections and something riskier: customers simply handing an AI tool their login credentials. Last month, Axios reported that major banks had detected third-party agents operating inside customer accounts — in some cases with the customers' permission and passwords.
Every bank email viewed by Business Insider told customers not to share usernames, passwords, or security codes.
Wells Fargo put it in policy language. "No technology should be trusted to act on a customer's behalf without strong security, clear customer authorization, and safeguards designed to protect their money and financial information," a bank representative told Business Insider.
What did Jamie Dimon say?
JPMorgan Chase CEO Jamie Dimon issued his own warning during an October earnings call.
"If people are using them, they should understand the risk," Dimon said. "In the meantime, we're not going to allow just any outside agent to come in inside the company because that's just too risky for clients and for JPMorgan."
That's the largest US bank saying it will actively bar outside agents from its systems — a signal of where the industry's tolerance currently sits.
Does the warning extend beyond banks?
Yes. The beta of Apple's latest Mac software includes warning language when a user considers granting an AI agent or app system-level access to their computer, according to MacRumors. Device makers are now flagging the same risk the banks are.
The message from every corner is consistent: the technology can act, but the liability stays with you. Until banks and regulators build formal safeguards for agent-initiated transactions, customers delegating their wallets to AI are doing it at their own risk.
Source: feeds.businessinsider.com
More from Grace Kim
Show full bio
Market editor covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.
2,387 articles