Incident sheet · /security

Filed
Read
3 min · 580 words
Severity
Routine

Citi shifts AI-agent hack losses to customers under new terms

Citi, JPMorgan, Bank of America and PNC have shifted third-party AI-agent risk to customers, who may owe funds even after credentials leak or an agent is hacked.

More banks says customers using AI agents are on their own — even if they get hacked
FeedMore banks says customers using AI agents are on their own — even if they get hacked — Erik_Lucatero / pixabay

Event log

  1. Citi, the third-largest U.S. commercial bank, updated its Online User Agreement this week.
  2. JPMorgan and Bank of America, the two largest U.S. banks, announced similar policies last month; PNC, the eighth-largest, changed its terms today.
  3. Citi says customers remain responsible for hacker activity after a third-party AI agent leaks credentials or is hacked.
  4. Anthropic launched a read-only personal finance tool earlier this month that cannot move, manage or invest money.
  5. Citizens emailed customers this week, warning that they bear responsibility for AI-assistant activity.

By Olivia Hart

Citi has shifted the financial risk of third-party AI agents to its customers, including losses after an agent leaks credentials or gets hacked.

A customer who hands an AI assistant the credentials for a bank account and other online financial portals may now own the outcome: the customer bears the cost of an intentional or unintentional transaction and any hacker activity that follows a breach.

Citi, the third-largest commercial bank in the United States, changed its terms this week through its Online User Agreement. Banks have grown wary of unreliable AI actors and say they will not reverse an intentional or unintentional transaction carried out by an automated assistant.

Third-party agents range from frontier laboratory models to purpose-built personal finance tools. Once connected, they can:

  • receive credentials to bank accounts and other online financial portals;
  • analyze spending and investment decisions;
  • offer financial advice; and
  • handle transactions and bill payments.

A Citi spokesperson said: “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 updated agreement, the spokesperson said, reinforces that customers who grant access remain responsible for activities conducted through their agents.

Banks widen the shift

  • JPMorgan and Bank of America, the two largest banks in the country, announced similar policies last month.
  • PNC, the eighth-largest bank, changed its terms today.
  • Citizens emailed customers this week, warning that they are on the hook for anything an AI assistant does, according to an email seen by TechCrunch.

Citi’s language goes further than transaction disputes. If a third-party AI tool leaks the customer’s credentials or gets hacked, the customer remains responsible for what hackers do with the account.

What does this change for customers?

For customers using third-party agents, a disputed payment may stay unreversed, while a theft enabled by leaked credentials may become the customer’s loss rather than the bank’s. In practical terms, connecting an agent transfers much of the operational and cybersecurity risk with it.

Citi’s terms say customers can use AI agents that Citi provides without additional liability. Anthropic took the narrower route: its personal finance tool, launched earlier this month, is explicitly billed as “read-only” and “can’t move, manage, or invest your money.”

Personal finance gives AI startups a route to new revenue and a share of the sector’s fee income. The pitch is to add value around customers’ accounts, but the trust barrier is higher than sharing an email or chat history. Consumers already have reservations about those records; bank and brokerage accounts require a new level of confidence.

Banks also view agent-led finance as a competitive threat. Markets shuddered briefly at the prospect of AI tools making it easier for consumers to avoid unnecessary fees or low-interest deposit accounts that pad bank margins. If agents become a primary route into financial decisions, banks will fight to control that channel.

The pattern extends beyond finance. AI labs promise agents that transform how customers interact with service providers, while those providers push back on responsibility for failures. The same tension already runs through car-hailing and food service, and its outcome will determine which AI investments pay off.

For banks, the winning agent may not be the one that can do the most; it may be the one that can act without leaving every customer carrying the loss.

via techcrunch.com (Source)

Filed under

  • ai-agents
  • cybersecurity
  • banking
  • risk-management
  • financial-services
Share this article:

More from Olivia Hart

Olivia Hart

Show full bio

News editor covering money, funding and market shifts around enterprise information technology and the leaders who run at CIO Monitor.

207 articles

The newsroom digest

One email a day with the stories that matter. No noise.

By subscribing you agree to the privacy policy.

Cross-references

‹ Previous articleNext article ›