Gave an AI Chatbot Your Bank Login? Citi Says You're Still on the Hook if It's Hacked

Citi, PNC and Citizens have joined Bank of America and JPMorgan in telling customers they're on the hook for whatever a third-party AI agent does.
We receive compensation from the products and services mentioned in this story, but the opinions are the author's own. Compensation may impact where offers appear. We have not included all available products or offers. Learn more about how we make money and our editorial policies.

If you've let an AI chatbot or assistant peek at your bank balance, check the fine print. Citibank quietly rewrote its customer agreement this week to say that if you hand a third-party AI tool your banking login and that tool gets hacked, the money a criminal drains from your account is your problem, not the bank's.[1]

Citi isn't alone. PNC changed its terms on November 18, and Citizens Bank emailed customers this week with a similar warning. JPMorgan and Bank of America already made comparable moves last month.

In about a month, four of the country's largest banks have told customers, in writing, that letting an AI agent into their account comes with a catch: you own whatever happens next.

In this article
What Citi actually changed
Why banks are moving now
What this means if you've connected an AI tool to your bank account

What Citi actually changed

Citi added a new section, Section H, to its Online User Agreement this week. It applies to anyone who lets a third-party AI agent access their accounts or digital services through Citi Online or Citi Mobile.

"As third-party AI tools become more widely available, we want customers to understand the risks associated with sharing banking credentials or authorizing those tools to access their accounts," the spokesperson said. "We recently updated our Online User Agreement to provide greater clarity about those risks and to reinforce that customers who choose to grant account access to third-party AI agents are responsible for activities conducted through those agents."

The practical effect goes well beyond a warning label. According to the language of Citi's updated terms, if you give your login credentials to a third-party AI agent, Citi will treat anything that agent does as something you authorized. If those same credentials later get shared, leaked, or stolen and end up in the hands of another person or another AI agent, Citi says it will still consider whatever happens with them as authorized by you. The bank states it has no obligation to investigate whether those actions were actually valid.

Citi's terms also put the burden of oversight on the customer, not the bank. You're responsible for mistakes the AI agent makes, including misreading your instructions, taking unintended actions, or going beyond the authority you gave it, except where the law says otherwise. Watching what the agent does is explicitly called "your responsibility and not Citi's."

Citi reserves the right to cut off any third-party AI agent's access at any time, without warning. By letting an agent into your account, you're also agreeing to let Citi share your personal information, including sensitive personal information, with that outside tool. Citi's own privacy notice doesn't apply to that agent, and the bank says it has no control over how the agent stores, protects, or shares your data.

A separate Citi agreement covering online funds transfers, updated with a November notice, goes a step further: it says any transfer an AI agent initiates on your behalf counts as authorized by you, even if it conflicts with what you actually told the agent to do.

Citi says it won't investigate those transactions further. One carve-out: AI tools Citi builds and offers itself, through Citi Online or Citi Mobile, are excluded from all of this. Use those, and you're not taking on the extra liability.

Why banks are moving now

This week's updates aren't the first shot. JPMorgan and Bank of America rolled out similar policies in October. On October 27, Bank of America told customers that using AI tools with their banking credentials means they're taking on the risk of whatever happens in that account.

The pattern is consistent across every bank that has acted: none of them is promising to reverse a transaction an AI agent made, whether you meant for it to happen or not. AI agents, whether built by frontier labs or purpose-made for personal finance, are increasingly given logins and asked to analyze spending, offer advice, pay bills, and move money.

Not every company building these tools wants that level of access. Anthropic's personal finance tool, launched earlier this month, is explicitly marketed as "read-only" and states it "can't move, manage, or invest your money." That's a notably more cautious design than agents built to act on your behalf with full account access.

Banks also have a competitive stake in how this plays out. If AI agents become the default way people compare rates, move money out of low-interest accounts, or avoid fees, that's a direct hit to bank margins. By controlling the terms around third-party access, banks are also drawing a line around who gets to build the tools that touch their customers' money.

What this means if you've connected an AI tool to your bank account

If you've already given a chatbot, browser extension, or finance app your banking login, it's worth checking your bank's current user agreement before you use it again. These aren't proposed changes. Citi, PNC and Citizens have already put them into effect, and courts haven't yet tested how far banks can push this kind of liability shift.

A few things worth doing now:

  • Pull up your bank's online user agreement or access agreement and look for new language about third-party AI agents. Several major banks have changed theirs in just the past month.
  • Avoid sharing your actual banking login and password with any outside AI tool. If you've already done it, consider revoking that access, since banks now hold customers responsible even if those credentials are leaked or stolen afterward.
  • If you want AI help with budgeting or spending analysis, look for tools that only need read access or are built directly into your bank's app, since those typically don't carry this extra liability.

The broader lesson echoes concerns already raised around AI shopping agents and other automated tools that act on your behalf: the more access you give a piece of software, the more you need to understand exactly what happens if that software fails, gets tricked, or gets hacked.

4.8
Editorial Rating
See Price
On Aura Identity Theft's website
2026 Editors’ Choice
Best Overall Identity Theft Protection Service
Identity Protection
Aura Identity Theft
  • ID theft protection that monitors your SSN, bank accounts, credit cards, and brokerage and retirement accounts for suspicious activity
  • Every plan includes the full feature set, so no additional cost to unlock monitoring, insurance, or restoration
  • Bundles data removal with identity theft protection, antivirus, VPN, and a password manager in one subscription

Author Details
Kalleigh Lane is the Managing Editor at All About Cookies, where she helped develop the site's star rating system for product reviews and specializes in testing VPNs, ad blockers, and parental controls on Android and PC. She brings more than five years of experience editing and writing across cybersecurity, tech, and finance, and she has authored and edited over 300 articles for All About Cookies. Before joining AAC, she worked as a journalist and editor at CTV News and the Globe and Mail, and she holds a Master of Arts in English Literature from the University of Toronto.

Citations

[1] More banks says customers using AI agents are on their own, even if they get hacked, TechCrunch