Executives at leading banks are weighing the risks of agentic AI as concerns over the technology’s swift development spread.
JPMorgan Chase CEO Jamie Dimon said during the bank’s Q3 2026 earnings call Tuesday that the bank is not “against agents or for agents,” pointing to the bank’s adoption of the technology both internally for employee use and externally for customers. However, he said the bank is not ignoring security concerns “because you don’t want rogue agents — and there could be thousands of them.”
As the bank works to build tools to protect customers, it will not allow all outside agents into the JPMorgan Chase environment, Dimon told investors.
“We’re not going to allow just any outside agent to come inside the company because that’s just too risky,” Dimon said.
While protecting customers in the agentic era is top of mind for JPMorgan Chase, the bank is also focused on minimizing lock-in by staying model-agnostic as a strategy to manage AI costs.
Dimon pointed to model routing tools that send queries to the quickest, cheapest models.
“You don’t need the most expensive thing for every query you have and very often people just see the one they’re comfortable with,” he said.
Managing AI costs, risks and benefits
Other major banks reporting earnings this week are also assessing the risks of AI agents, outlining the benefits they’re seeing and identifying strategies for managing costs.
At Wells Fargo, CEO Charlie Scharf said the bank is excited for the opportunities AI presents to both the company and customers, but is also “taking a disciplined approach to managing risk.”
“We will also evaluate how third-party models and agents can provide value-added capabilities to our clients, while at the same time considering the potential risk these tools can introduce for both our clients and Wells Fargo,” he said during the company’s Q3 2026 earnings call Tuesday.
Bank of America continues to view AI as a driver of franchise growth, CEO Brian Moynihan said during Wednesday’s Q3 2026 earnings call. The bank is pushing ahead with automating processes and simplifying workflows, meaning less hiring to refill roles, executives said during the call.
On Tuesday, the bank was fined $1.2 billion for violating anti-money laundering and sanctions laws — the third-largest fine ever levied by the Treasury Department.
While aware of AI risks, Citi CEO Jane Fraser said during the bank’s third quarter earnings call Tuesday that it’s also seeing numerous benefits driven by adoption of the technology.
Fraser said AI is enabling the bank to launch new products faster and at a lower cost, seeing an average 30% to 40% increase in product launch speed. Meanwhile, 14,000 of the bank’s engineers use its main agentic coding tool Devin to make in-house work more efficient. The bank is also reengineering 100 manual processes using AI.
“We’ve been in-sourcing work,” she said during the call. “We’ve reduced our tech vendor spend by 30% year-over-year as a consequence, and that’s still got a long way to run.”
At Morgan Stanley, CEO Ted Pick described AI’s power as an efficiency, productivity and resiliency tool as incontrovertible.
The technology is “fundamentally changing what our teams do,” said Morgan Stanley CFO Sharon Yeshaya.
“Our investments in technology and talent have strengthened our capabilities, allowing us to serve at greater scale and manage risk more effectively in fast-moving markets,” she said during the bank’s Q3 2026 earnings call Wednesday.
