AI··2 min readTechnical readSource-backed

Banks Move the Blame for AI‑Powered Finance Into Customers’ Hands

Citibank just told its customers that if an AI assistant hacks their account, the blame is on them.

By
Editorial illustration for Banks Move the Blame for AI‑Powered Finance Into Customers’ Hands
WireUnwired deterministic editorial visual
WIREUNWIRED BRIEF

Citibank just told its customers that if an AI assistant hacks their account, the blame is on them. The bank’s new terms of service make it clear that anyone who gives a third‑party AI tool access to their banking credentials is responsible for every action that tool takes, even if the tool is compromised.

AI agents can read account data, analyze spending, and even move money, but banks are wary of the uncertainty they bring. The new language follows a wave of similar moves by JPMorgan, Bank of America and PNC. Citi’s spokesperson said the update “provides greater clarity about the risks associated with sharing banking credentials or authorizing those tools to access your accounts.” The bank also warned that if an AI tool leaks credentials or is hacked, customers will be liable for any resulting theft.

Startups see personal finance as a goldmine, but the trust hurdle is steep. Anthropic’s read‑only tool, for example, is billed as “can’t move, manage, or invest your money.” Yet banks view the rise of AI assistants as a competitive threat that could let customers sidestep fees and low‑interest accounts.

Our read: the shift in liability is a sign that banks will fight to control the AI pathway.

As AI labs promise to change how we interact with services, the real battle will be who owns the conversation and the money.

Watch how banks tighten terms and whether AI tools can still offer a safe, frictionless experience for everyday banking.

STORY JOURNEY

The story so far

Earlier coverage connected by shared names and topics. Follow the dates to see how coverage developed; these links do not imply every article describes the same event.

  1. Editorial illustration for Walmart Introduces New Agent Gateway to Allow Approved Artificial Intelligence Shopping Assistants for Customers
    Walmart Introduces New Agent Gateway to Allow Approved Artificial Intelligence Shopping Assistants for Customers

    Walmart announced a new Agent Gateway to manage transactions and features for customers using artificial intelligence agents. The move contrasts with competitor…

    Connected by: CustomersEARLIER COVERAGE
  2. Editorial illustration for Comcast Loses 144,000 Broadband Customers in Q3 2026 and Plans Major Restructuring Overhual
    Comcast Loses 144,000 Broadband Customers in Q3 2026 and Plans Major Restructuring Overhual

    Comcast reported a net loss of 144,000 residential broadband customers in the third quarter of 2026, up from a 91,000-customer loss during…

    Connected by: CustomersEARLIER COVERAGE
  3. Editorial illustration for Bank of America Asks You Not to Hand Your Bank Login to an AI Bot
    Bank of America Asks You Not to Hand Your Bank Login to an AI Bot

    Financial giant Bank of America recently sent an email blast to customers urging them never to share their usernames and passwords with…

    Connected by: CustomersEARLIER COVERAGE
  4. Editorial illustration for Unsiloed AI Raises $7.5M to Turn Messy Business Documents Into Fuel for AI Agents
    Unsiloed AI Raises $7.5M to Turn Messy Business Documents Into Fuel for AI Agents

    AI agents are hitting a wall of messy paperwork. While companies rush to automate tasks, critical information remains trapped inside complex layouts,…

    Connected by: IntoEARLIER COVERAGE
  5. NOW
    Banks Move the Blame for AI‑Powered Finance Into Customers’ HandsCURRENT UPDATE
A WireUnwired summary is already available. You can still write your own brief below for the reader community.
SOURCE & ATTRIBUTION
TechCrunch

Original author: Tim Fernholz

Original publication: 18 November 2026 23:02

Read full story on TechCrunch ↗
RELATED STORIES
Editorial illustration for OpenAI Unveils AI‑Powered CSAM Detector That Cuts False Positives in HalfTechCrunchOpenAI Unveils AI‑Powered CSAM Detector That Cuts False Positives in HalfOpenAI chip logo graphic representing the article on California suing OpenAI over autonomous AI incidents.Ars TechnicaCalifornia Sues OpenAI Over 80 Autonomous AI IncidentsGoogle Gemini branding and interface graphicArs TechnicaGoogle Finally Lets You Keep Your Gemini History Without Donating Your Data
Original sourceTechCrunch ↗