As artificial intelligence-powered personal assistants gain traction, JPMorgan Chase is drawing a clear line on how far third-party AI applications can go when it comes to accessing customer accounts.
During the bank's third-quarter earnings call, executives addressed growing investor concerns around "agentic AI" and the rise of AI assistants that can act on behalf of users. While the lender sees significant potential in the technology and is already deploying AI agents internally, it made clear that customer protection remains its top priority.
"We are not gonna allow just any agent, outside agent, to come inside the company because it is just too risky for clients and for JPMorgan," CEO Jamie Dimon said.
The comments come as technology companies roll out increasingly sophisticated AI assistants capable of carrying out tasks such as managing finances, comparing products and, potentially, initiating transactions on behalf of users. Such tools have sparked debate over whether consumers could eventually rely on AI intermediaries instead of interacting directly with banks.
JPMorgan does not view the emergence of AI agents as a completely new competitive threat. According to Chief Financial Officer Jeremy Barnum, the trend is largely an extension of two existing challenges banks have dealt with for years: competition for deposits and competition from fintech firms offering better digital experiences.
However, the bank believes the risks increase when customers grant third-party applications direct access to their accounts.
Barnum specifically highlighted concerns around credential sharing, where users provide usernames and passwords to external applications. He noted that some services continue to operate through methods such as screen scraping, which JPMorgan has long opposed on security grounds.
"It is not secure," Barnum said. He added that customers should understand the potential consequences if an externally controlled AI agent is given account credentials and then performs actions the user did not intend.
The discussion sheds light on what consumers using AI-powered financial tools may need to consider. While AI assistants can help compare products, analyse spending or provide recommendations, banks remain wary of allowing autonomous systems to directly move money or execute transactions without strong safeguards in place.
JPMorgan said it is actively developing its own tools to help protect customers and manage these emerging risks. At the same time, executives stressed that the bank is embracing AI rather than resisting it.
Dimon said the lender already uses AI agents internally and will continue to expand those applications. Barnum also pointed to growing productivity gains from generative AI, particularly in software engineering and internal operations, although he noted that those benefits are not yet materially visible in the bank's earnings.
Beyond technology, executives argued that customer trust remains one of JPMorgan's biggest competitive advantages in an era of AI-driven finance. Barnum said the bank's longstanding investments in its branch network, digital capabilities and customer relationships position it well as the technology evolves.
Still, he cautioned that trust alone is not enough.
"The fact that we are trusted is not an excuse for complacency," Barnum said, adding that banks must continue improving customer experiences even as new AI-powered competitors emerge.
For consumers, the message is straightforward: AI assistants are becoming more capable, but banks are likely to remain cautious about giving third-party agents unrestricted access to customer accounts until security, liability and consumer protection concerns are fully addressed.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.