JPMorgan blocks external AI agents from accessing customer accounts

By Splitfeed EditorialLast updated Oct 13, 2026, 22:20 UTC

On October 13, 2026, JPMorgan Chase CEO Jamie Dimon publicly confirmed that the bank has blocked all external AI agents from accessing customer accounts. Dimon stated the bank is not using any outside AI agents and described the restrictions as protective gates erected specifically against what he termed "rogue agents." The announcement was reported across financial and technology outlets the same day, with Dimon's framing centering on risk management rather than a wholesale rejection of AI technology.

The move carries weight beyond a single corporate policy decision. JPMorgan Chase is among the largest financial institutions in the world, and a public statement from its CEO sets a visible precedent at a moment when AI agents, meaning autonomous systems capable of taking actions on behalf of users, are being integrated into consumer-facing financial tools by a range of companies. The core dispute is whether blocking external agents is a prudent and necessary safety measure or an overcautious reaction that will slow the adoption of genuinely useful technology. What remains unspecified is which particular agent systems or platforms are caught by the restriction, leaving the practical scope of the policy open to interpretation.

With no published reactions yet from the Pro-AI, Anti-AI, or Middle Ground camps, their likely positions can be anticipated based on how each typically approaches stories of this kind. Pro-AI voices would probably argue that blanket restrictions stifle innovation and that the risks Dimon describes are manageable through better standards and oversight rather than outright exclusion. Anti-AI voices would likely welcome the move as an overdue acknowledgment that autonomous systems interacting with sensitive financial data pose serious and underappreciated dangers. Middle Ground commentators would typically call for clearer regulatory frameworks that distinguish between trustworthy and risky agent architectures, arguing that a binary block-or-allow approach is too blunt for the complexity involved.

The argument is likely to sharpen as other major banks respond publicly or quietly follow suit. Any regulatory guidance from bodies like the Office of the Comptroller of the Currency or the Consumer Financial Protection Bureau on AI agent access to financial accounts would do a great deal to settle whether JPMorgan's position becomes an industry standard or an outlier.