Breaking
US'Thursday Night Football' prediction, odds, spread, time: Colts vs. Texans picks from expert on a 34-20 runFREtats-Unis : De passage à Paris, Kamala Harris va rencontrer Emmanuel MacronITX Factor 2026, nella notte degli inediti eliminata camilla. DIRETTACN美国航空航天局:收到探测器从距离地球一光日发来的信号CN尼泊尔法院批准羁押两名前劳工部长CNAnthropic上市 创史上最大IPO融资纪录INTLJapan reports cooler-than-expected core inflation in October of 2.2% — but highest this year so farITGenova, aeroporto chiuso per due ore per sospetta presenza di droniUKAnthropic becomes one of world's most valuable public companies in stock market debutJP【そもそも解説】復元された首里城とは? 沖縄そのものと重なる歴史US'Thursday Night Football' prediction, odds, spread, time: Colts vs. Texans picks from expert on a 34-20 runFREtats-Unis : De passage à Paris, Kamala Harris va rencontrer Emmanuel MacronITX Factor 2026, nella notte degli inediti eliminata camilla. DIRETTACN美国航空航天局:收到探测器从距离地球一光日发来的信号CN尼泊尔法院批准羁押两名前劳工部长CNAnthropic上市 创史上最大IPO融资纪录INTLJapan reports cooler-than-expected core inflation in October of 2.2% — but highest this year so farITGenova, aeroporto chiuso per due ore per sospetta presenza di droniUKAnthropic becomes one of world's most valuable public companies in stock market debutJP【そもそも解説】復元された首里城とは? 沖縄そのものと重なる歴史
BackCiti Says Customers Are Responsible When AI Banking Agents Go Rogue—Even After a Hack
Citi Says Customers Are Responsible When AI Banking Agents Go Rogue—Even After a Hack
Developing
Decrypt1 hour agoBusiness2 min read

Citi Says Customers Are Responsible When AI Banking Agents Go Rogue—Even After a Hack

Major banks shift responsibility to customers when outside AI agents access accounts, even if credentials are leaked or misused

Quick Look

  • Citi has updated its online banking terms to make customers responsible for actions taken by third-party AI agents accessing their accounts, including credential leaks and unauthorized access.
  • JPMorgan, Bank of America, Citizens Bank, and PNC have announced similar policies.
  • Federal consumer protections still apply, though their application to AI-authorized transactions remains unclear.

AI-generated summary

Why It Matters

Citi revised its online banking terms in November to clarify that customer-supplied credentials to outside AI agents are treated as customer-authorized actions. JPMorgan and Bank of America announced similar policies in October, Citizens Bank warned customers by email this week, and PNC changed its digital agreement on Wednesday. The OCC issued nonbinding AI examiner guidance addressing outside agents, payment controls, and human approval. The CFPB states banks cannot use agreements to waive federal consumer fraud protections.

Font size

Citi has updated its online banking terms to make customers responsible for the actions of third-party AI agents accessing their accounts, including when an agent leaks their credentials and someone else uses them to get in.

The change, reported by TechCrunch Wednesday, makes explicit how the bank intends to handle an increasingly awkward question: Who pays when software handling your money makes a mistake?

Under Citi's new terms, actions taken using credentials supplied to an outside agent are treated as authorized by the customer, even if the tool ignores instructions, exceeds its permissions, or shares those credentials with another party.

Citi says it has no obligation to investigate whether the agent followed the customer's original prompts, subject to applicable law.

A Citi spokesperson told TechCrunch that the November revision clarifies the risks of sharing banking credentials or granting outside tools account access. The provision does not apply to AI agents offered by Citi itself.

Banks tighten the rules on AI access

Citi is not acting alone. JPMorgan and Bank of America announced similar policies in October, while Citizens Bank warned customers this week by email that they could be responsible for what their AI assistants do.

PNC also changed its digital agreement on Wednesday. Its rules distinguish bank-approved connections, which can provide limited access, from handing an agent unrestricted access to an account.

The move follows a banking security scare from October 22. Major banks had detected AI agents logging into customer accounts, sometimes taking steps that appeared to go beyond their users' instructions, and executives said they struggled to determine which agents had been properly authorized.

An AI agent differs from a conventional chatbot because it can act on a user's behalf, potentially navigating websites, accessing financial records, or initiating payments. That makes a mistaken instruction or compromised login more consequential than a bad answer about a bank balance.

The Office of the Comptroller of the Currency has since issued AI examiner guidance that addresses outside agents, payment controls, and human approval. The framework is nonbinding and does not itself create new requirements for banks.

A banking agreement cannot erase federal protections

Citi's language does not automatically eliminate consumer fraud protections. Under U.S. federal consumer protection rules, banks generally must investigate unauthorized electronic transfers and limit customers' financial liability for fraud. Those protections cannot simply be waived through a banking agreement, although their application to transactions performed by customer-authorized AI agents remains less clear.

The Consumer Financial Protection Bureau says banks cannot use agreements to waive those legal protections, although the rules distinguish theft by an outsider from transfers actually authorized by a customer or someone given access.

Whether a particular AI-related loss qualifies as unauthorized depends on how the transaction occurred.

PNC's updated agreement takes a stricter operational approach: Customers may use certain connection methods approved by the bank, but cannot give third-party AI agents their passwords or one-time security codes, and agents generally cannot initiate payments unless PNC expressly permits it.

What to Watch

AI outlook — possibilities, not facts

  • The CFPB or OCC may issue further guidance or rules clarifying liability for AI-authorized banking transactions

    Likely · Within months

  • More banks will adopt similar terms shifting liability to customers for third-party AI agent actions

    Likely · Within months

Open Questions

  • How will banks distinguish between customer-authorized AI agent transactions and truly unauthorized fraud?
  • What specific steps must customers take to remain protected under federal fraud rules when using AI agents?
  • Will the CFPB or OCC issue binding rules on AI agent liability in banking?
  • How many customers are currently using third-party AI agents to access bank accounts?

Related Topics

This article was originally published by Decrypt.

Related Stories

Bullish recruits Wall Street heavyweights to solve tokenized stocks’ biggest trading problem
Developing·

Bullish recruits Wall Street heavyweights to solve tokenized stocks’ biggest trading problem

Bullish and Equiniti announced that DRW, Flow Traders, SoFi and Superstate have joined the Issuer Token Coalition, bringing membership to 12 companies including the New York Stock Exchange. The group is developing common standards to make tokenized stocks tradeable across platforms, addressing barriers that currently confine tokenized shares to issuer-specific arrangements.

CryptoSlate
3 min read
More on this topicciti