Major Banks Alarmed by Unauthorized AI Agents in Customer Accounts
Major banks report unauthorized third-party AI agents accessing customer accounts, raising security concerns.
Several major U.S. banks are reporting unauthorized activities linked to third-party AI agents accessing customer accounts without permission. This issue involves customers sharing their login credentials with AI tools that have not been vetted by the banks, according to a recent Axios report.
Madison Mills, a senior AI reporter for Axios, explained that many customers are unwittingly providing their usernames and passwords to these AI systems, which goes against bank policies. She pointed out instances where agents acted on behalf of account holders without their explicit approval, including making contact with banks over complaints.
Mills clarified that these agents are not hacking accounts or depleting funds. Instead, they are carrying out user instructions without the bank’s consent to perform transactions. “It’s really a nightmare of AI agent activity for banks,” she remarked.
David Sacks, the White House AI czar, emphasized the difference between user-authorized AI agents and malicious actors, stating that the actions described in the report stem from customers willingly sharing their information. He noted that while this practice breaches bank regulations requiring human logins, it does not equate to cyberattacks.
Sacks suggested that banks should accommodate these AI agents by providing them with “read-only” access to customer accounts. This would allow them to analyze account spending without making unauthorized transactions.
The growing issue has prompted banks to adapt quickly to prevent AI agents from accessing customer data. A Chase spokesperson advised customers against sharing their credentials and emphasized the importance of secure, authorized data sharing through open banking providers.
Concerns were echoed by JPMorgan Chase CEO Jamie Dimon during an earnings call, where he stressed the risks associated with allowing third-party agents into customer accounts. He stated, “We’re not going to allow just any outside agents to come inside the company because it’s just too risky for clients and for JPMorgan.”
In response to the rising threat, Charles Schwab’s CEO confirmed that the company is actively blocking unauthorized AI agents. However, Morgan Stanley reported no instances of unauthorized agent activity across its platforms. Bank of America and Wells Fargo did not provide immediate comments on the situation.
