AI

Major US Banks Warn Customers Using AI Finance Agents They Bear Full Risk, Even If Hacked

Tendela Briefing ·
8 views · 0 Comments
More banks say customers using AI agents are on their own — even if they get hacked | TechCrunch

Image: TechCrunch · Source

Citibank, along with JPMorgan, Bank of America, and PNC, has updated terms to shift responsibility onto customers using AI personal finance tools, including risks from hacking or unauthorized transactions.

Citibank, the third largest U.S. commercial bank, recently revised its Online User Agreement to clarify that customers who authorize third-party AI agents to access their accounts are solely responsible for any resulting activity — including unauthorized transactions and hacks. This follows similar policy changes from JPMorgan, Bank of America, and PNC, signaling a broader industry stance in the wake of growing use of AI personal finance assistants.

AI agents are software tools that can be granted banking and brokerage credentials to analyze spending, provide financial advice, automate transactions, or manage bill payments. While these capabilities promise convenience and enhanced financial insights, banks have expressed concerns about security and control. They emphasize that since these AI tools operate outside their direct supervision, customers carry the liability for misuse or errors.

Citibank’s updated terms explicitly state that if a customer’s AI agent leaks credentials or is compromised by hackers, the customer—not the bank—is responsible for any resulting losses. Similarly, Citizens Bank recently notified its clients of the same obligations. This policy shift comes as AI startups eye personal finance as a lucrative sector, aiming to monetize through fees tied to enhanced services and account management.

Some AI providers attempt to address trust issues; for example, Anthropic’s new personal finance tool highlights a "read-only" mode, preventing the AI from moving or investing money, reducing risk exposure. Still, banks view independent AI tools accessing accounts as potential competitive threats, worrying that easier fee avoidance and lower-margin account adoption could erode profits.

The banks’ insistence on user accountability is part of a wider trend across industries responding to the rise of AI agents. Many service providers are cautioning customers about entrusting their credentials or relying on external AI agents, highlighting the evolving balance between AI innovation and consumer protection. How these dynamics unfold will heavily influence which AI-driven financial products succeed and gain mainstream adoption.

In contrast, banks’ own AI implementations come with assurances of limited liability, underscoring their desire to maintain tighter control over AI interactions within their ecosystems. Customers considering third-party AI financial assistants should carefully weigh convenience against the potential risks and read updated user agreements closely.

Sources and original reporting

Comments (0)

No comments yet. Start the discussion.

Write a comment

Comments are published after moderation. Your name and comment will be visible publicly. Account