AI agents are advancing faster than financial institutions can keep up, raising concerns about their access to consumers’ bank accounts.
American banks are working to restrict third-party AI agents accessing their platforms, citing concerns over how the technology interacts with customer accounts, Axios reports. Morgan Stanley, Bank of America, JPMorgan Chase, Citi, and Wells Fargo are among the institutions working to contain the risks.
Inside Financial Institutions’ Decision To Restrict AI Agents’ Access To Customer Bank Accounts
As AI capabilities advance, banks face the challenge of distinguishing between authorized activity and actions taken without a user’s direction, per Axios. Even when users allow agents to log in with their credentials, the activity can violate the policies of most banks that require customers to access their accounts themselves, the outlet notes.
Bank executives have reportedly described the risks as alarming, expressing serious concerns about AI agents logging in to customer accounts and taking actions that could affect users’ finances. Institutions have already noted that some AI agents have appeared to operate without a user’s express instructions. A Chase spokesperson cautioned customers against sharing usernames and passwords, underscoring security concerns about third-party access.
Rather than granting agents full access to banking functions, many banks have authorized, or are moving toward, read-only access. The move enables AI agents to review customers’ financial information without making purchases, transferring funds, or changing account activity.
Financial institutions are now developing strategies to prevent worst-case scenarios while preparing for the possibility that they could occur. According to Axios, preparation includes “day after” scenario planning to help banks quickly secure their systems and restore customer access following a potential incident.
While banks are in a solid position to respond to AI’s risks and opportunities, one source cautioned that having certain financial resources may not be enough as the technology continues to evolve and expose new vulnerabilities in real time.
As AI Agents Advance, Concerns Extend Beyond Banking
The risks surrounding AI extend beyond banking as AI agents increasingly take on tasks traditionally handled by humans. In the literary world, several AI agents developed by users on iLands have reportedly begun pitching literary magazines for work as writers without human oversight, AFROTECH™ previously reported.
As AI tools become more autonomous, industries must decide how much authority to grant AI systems — and what happens when those systems take actions their users did not intend.
Jacob Coxon, a former OpenAI and Anthropic researcher, has also raised concerns about the technology’s trajectory, warning that some people developing AI believe it could “kill us all by the end of the decade,” according to a previous report by AFROTECH™.
“Do not underestimate the power of this technology,” Coxon wrote on X on Sept. 8, 2026. “These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing.”
He added, “This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible – but I hear the same people express fear privately. No other human activity poses this level of danger.”
As AI systems become increasingly capable of operating with limited human intervention, questions about their oversight, security, and potential consequences are becoming more urgent. From financial institutions protecting customer accounts to AI agents independently carrying out tasks across industries, the challenge is no longer just what the technology can do, but how to ensure it operates within boundaries that protect its users and the world at large.