The bank to deploy more powerful agents this year

JPMorgan Chase CNBC plans to deploy AI agents later this year that can operate autonomously for a much longer period than current versions, marking another milestone in enterprise adoption of AI, CNBC has learned exclusively.

AI agents are evolving from tools that complete single tasks to digital workers that manage workflows through multiple steps and disparate software, Derek Waldron, chief analytics officer at JPMorgan, told CNBC in an interview.

“We have now entered the age of longstanding independent agents,” Waldron said. This “means that agents are not just running for two or three minutes to execute a goal or some human instruction, but rather they can run for an hour or two.”

Long-running clients have already emerged over the past year with examples including Anthropic’s Claude Code and OpenClaw going viral. However, JPMorgan’s planned deployment indicates that the technology is close to removing the security and governance hurdles that have slowed its adoption within large companies.

JPMorgan, run by CEO Jamie Dimon since 2006, is the largest U.S. bank by assets and has an annual technology budget of about $20 billion.

While much of the conversation about generative AI has focused on modular intelligence, technology leaders are increasingly focusing on a different question, Waldron said: How long can AI systems operate effectively before they require human intervention?

This concept, which Waldron called “intellectual coherence,” has been helped by improvements in how AI models think, enabling them to be “more team managers than individual workers,” he said.

“Just like the way people work, team managers can analyze the problem and delegate activities, and teams can work much longer to do more complex things,” Waldron said.

Other recent advances that have helped agents perform more complex tasks include the ability to write code, control web browsers, and interact directly with desktop programs, he said.

While the long-running agents are not yet ready for corporate use due to security concerns, their arrival is not far off. “We will have those in 2026,” Waldron said.

Ultimately, the AI ​​agents will remain coherent “for hours, then days, then weeks,” he said.

AI-driven productivity gains have been most evident in software development and back-office operations, but Waldron said it is increasingly enhancing revenue-generating roles.

In private banking, for example, AI systems scan market activity, customer attitudes and research overnight, helping bankers focus on customer interactions.

He said the bank has seen a 20% increase in overall sales due to these tools, and believes they could eventually allow retail bankers to expand client coverage by up to 50%.

Dimon has been clear that some of his workers will be replaced by artificial intelligence, saying the company is preparing to train and redeploy employees affected by the changes.

But Waldron added that while many companies initially approached AI as a cost-cutting tool, they are increasingly recognizing its potential to increase revenue.

“For companies to win with AI, it’s not about cutting the maximum number of jobs,” he said. “It’s all about trying to create a sustainable competitive advantage.”

The bank’s thinking about building software from outside vendors versus buying it has also changed, Waldron said. JPMorgan is now looking closely at whether it can build internal capabilities, which could put pressure on some traditional sellers, he said.

“Certainly the moat around certain types of software companies has narrowed compared to what it was in the past,” he said.

— CNBC’s Gabrielle Vonrughe contributed to this report.