JPMorgan Blocks Outside AI Agents From Customer Accounts, Dimon Says

Speaking on the bank's third-quarter earnings call, Dimon said JPMorgan would not allow just any outside agent to come inside the company because it is too risky for clients and for JPMorgan. "You don't want rogue agents, and it could be thousands of them," he told analysts, describing the potential threat to customers.

A JPMorgan spokeswoman confirmed to CNBC that the bank currently blocks external AI agents from operating on behalf of retail customers. "Protecting our customers is our top priority. Because it is not safe, we do not permit third parties to access accounts by obtaining customers' banking credentials and logging in on their behalf," she said. She added that the bank supports secure, customer-authorized data sharing through open banking providers and is committed to enabling agentic connections securely.

The restriction means AI agents can reach some Chase account data in a read-only fashion, such as customer balances, if a connector such as Plaid holds the user's credentials. They cannot send money or pay a bill, the actions that make an agent useful. The spokeswoman said there is no timeline for when safeguards would be in place to allow agents to interact directly with customer accounts.

Dimon said JPMorgan is using AI agents internally and plans to deploy them for customer-facing applications. He framed the bank's approach as neither for nor against the technology, but as a matter of ensuring it serves customers safely. Chief Financial Officer Jeremy Barnum said Tuesday that JPMorgan has begun a targeted pilot for its own AI tool, Smart Cash, which aims to predict a user's cash flow and move funds to maximize yield. According to CNBC, the tool appears to be a defensive move against external agents.

Dimon also drew a line on liability when customers authorize outside AI agents to act for them. "If people are using them, they should understand the risk," he said. "They should understand there's a liability transfer."

The position from Dimon, who leads the largest U.S. bank by assets and is a frequent spokesman for the U.S. financial sector, echoes an earlier fight with fintech firms over access to customer data, when JPMorgan moved to charge data middlemen including Plaid for access. CNBC noted that in July an OpenAI model hacked into AI company Hugging Face's servers.

Personal agents have gained mainstream popularity in recent months as AI models integrate more autonomy, including the ability to cancel subscriptions or use credit cards on a user's behalf. Instinct, a digital concierge that relies primarily on text messaging instead of an app, launched earlier this year on an invitation-only basis; the San Francisco startup behind it, Spear Street Technology, was founded last year and reached a $10 billion valuation last month. In September, bank stocks fell after Meta's Muse agent topped app download charts shortly after it was unveiled, sparking fears that consumers would use agents to move funds from low-yielding accounts to higher-yielding ones. Weeks later, OpenAI released its own agent, Dots. Apollo chief economist Torsten Slok and others have pointed to the risks that agents pose to cheap bank deposits, a risk that remains limited if agents are not allowed to operate freely with customer accounts.