Major US banks block AI agents, Sparking Policy Clash with White House

David Sacks urges compromise as banks detect unauthorized AI access in customer accounts

By LineZotpaper

Published23 October 2026

Read Time2 min

A new Axios report reveals major US banks are detecting and restricting personal AI agents operating within customer bank accounts. The institutions argue these unvetted tools, even when authorized by the account holder, violate security policies. The White House AI czar has distinguished the activity from cyberattacks and recommended a compromise that grants agents read-only access.

Major US banks are scrambling to block personal AI agents accessing customer accounts, according to a new report from Axios. Madison Mills, senior AI reporter for Axios, told CBS News that the phenomenon is widespread. In many cases, customers share their login credentials with an AI agent the bank has not vetted. Mills said evidence shows agents sometimes take actions users did not explicitly authorize, such as calling a bank to file a complaint. She stressed the behavior is not hacking, but agents acting on user instructions outside bank permission.

White House AI czar David Sacks responded on X, drawing a sharp line between user-authorized agents and cyberattacks. 'There's a world of difference between user-authorized agents (Muse, Instinct, Grokbot, etc.) and true unauthorized/rogue agents that could constitute cyberattacks,' Sacks wrote. He suggested banks offer read-only access, letting agents view data like spending habits without initiating transactions. 'Full write access (with user credentials) is a different risk category and should be approached carefully,' he added.

Banks are responding firmly. JPMorgan Chase CEO Jamie Dimon said on an earnings call the bank would not allow unvetted outside agents onto its systems. A Chase spokesperson told CBS News that customers should not share passwords. Charles Schwab's CEO said the company is blocking unauthorized agents. Morgan Stanley reported it had not observed the activity, while Bank of America and Wells Fargo did not respond to a request for comment.

Analysis

Why This Matters

The outcome will define whether consumers can use autonomous AI agents to manage their finances.
It sets a critical precedent for AI agency versus institutional security across regulated industries.
The clash directly challenges the decades-old model of personal authentication and liability.

Background

The rapid adoption of personal AI agents has outpaced security frameworks in many sectors. Banks, bound by strict regulations and liability for fraud, require human-level authentication for account access. AI agents inherently break this model, needing credentials to operate on a user's behalf.

Key Perspectives

[Banks]: Emphasize security and liability. Unauthorized agents violate terms of service and expose clients and institutions to risk they cannot control. [AI Advocates/White House]: Advocate for user agency over data. David Sacks argues that an AI agent authorized by a user is the digital equivalent of a personal assistant, not a breach, and permission tiers can manage risks. [Critics/Skeptics]: Question whether users understand the full scope of permissions they grant and the potential for agents to act beyond defined instructions.

What to Watch

Whether major banks adopt the read-only permission model suggested by the White House.
Regulatory guidance from agencies like the Consumer Financial Protection Bureau on AI agent access.
Technical measures banks deploy to detect and block AI agents versus developers designing agents to mimic human users.

Sources

AI agents detected inside customer bank accounts, report says · Moneywatch - CBSNews.com

Checked

11 claims in this story checked against 1 source: 8 supported.

Show each claim

Partly supportedMajor US banks are blocking personal AI agents from customer accounts.Moneywatch - CBSNews.com: “Mills noted that it's an industrywide problem facing dozens of financial institutions as they race to curb AI agents' access to customer accounts.”
Partly supportedBanks argue that unvetted AI agents violate security policies even when customers authorize them.Moneywatch - CBSNews.com: “In a lot of cases, this is just going to be customers giving their username or password to an AI agent that the bank hasn't had a chance to vet yet, that goes against the bank's terms of service,”
SupportedDavid Sacks urged a compromise allowing agents read-only access.Moneywatch - CBSNews.com: “He believes banks should support the use of such user-authorized AI agents, but that if banks want to limit their access, they could grant AI agents "read-only" permission,”
SupportedMadison Mills told CBS News the issue is widespread.Moneywatch - CBSNews.com: “Mills noted that it's an industrywide problem facing dozens of financial institutions as they race to curb AI agents' access to customer accounts.”
SupportedCustomers often share their login credentials with an AI agent the bank has not vetted.Moneywatch - CBSNews.com: “In a lot of cases, this is just going to be customers giving their username or password to an AI agent that the bank hasn't had a chance to vet yet,”
SupportedAgents sometimes take actions customers did not authorize, including calling a bank with a complaint without consulting the customer.Moneywatch - CBSNews.com: “Examples of agent behavior that a user never approved include the AI calling a bank with a complaint without first consulting the account holder.”
SupportedSacks wrote: “There's a world of difference between user-authorized agents (Muse, Instinct, Grokbot, etc.) and true unauthorized/rogue agents that could constitute cyberattacks,”Moneywatch - CBSNews.com: “"There's a world of difference between user-authorized agents (Muse, Instinct, Grokbot, etc.) and true unauthorized/rogue agents that could constitute cyberattacks," he said.”
SupportedSacks added: “Full write access (with user credentials) is a different risk category and should be approached carefully,”Moneywatch - CBSNews.com: “"Full write access (with user credentials) is a different risk category and should be approached carefully," he said.”
Partly supportedJamie Dimon said JPMorgan Chase would not allow unvetted outside agents onto its systems.Moneywatch - CBSNews.com: “"We're not going to allow just any outside agents to come in inside the company because it's just too risky for clients and for JPMorgan," he said.”
SupportedA Chase spokesperson said customers should not share passwords.Moneywatch - CBSNews.com: “"Customers should not give their username and passwords to anyone. We support secure, customer-authorized data sharing through open banking providers to meet customers' needs while keeping their accounts safe," a Chase spokesperson told CBS News.”
SupportedCharles Schwab's CEO said the company is blocking unauthorized agents.Moneywatch - CBSNews.com: “Charles Schwab's CEO earlier this month said the company has been blocking unauthorized agents, according to Reuters.”