AI agents move from experiment to enterprise reality, executives say
AI agents are no longer a speculative experiment inside corporate America. According to technology leaders gathered at The Wall Street Journal’s Technology Council Summit in Silicon Valley, the question is no longer whether AI agents belong in large enterprises, but how quickly they can be deployed safely and at scale.
A year ago, summit attendees were still debating whether autonomous AI bots had found their place in enterprise systems. This year, the tone has shifted decisively. AI agents are now being embedded across departments, from customer service to financial operations and software development.
Yet while adoption is accelerating, executives acknowledge that governance, cybersecurity and workforce concerns remain unresolved.
From pilot programs to daily use
Kathy Kay, chief information officer and executive vice president at Principal Financial Group, said enterprise AI usage has changed dramatically in just twelve months.
“You can’t do anything hardly anymore without your employees using AI,” Kay said. “To really get the pace and scale and things that you need, you have to use AI to get there.”
The shift is not limited to Silicon Valley. Principal Financial Group, founded in 1879 and headquartered in Des Moines, Iowa, is far removed from the tech startup ecosystem. Yet like many legacy firms, it is integrating AI tools across workflows.
At Sonic Automotive, chief technology officer Stephen Carvelli said AI agents are enhancing customer engagement.
“You can, with agents, be there 24, 7,” Carvelli said, describing around the clock service capabilities that would be costly or impractical using only human teams.
Similarly, Bank of New York Mellon has deployed 130 AI powered “digital employees,” each assigned a human manager. The structure reflects a hybrid model, where automation is overseen rather than fully autonomous.
Coding revolution accelerates
In Silicon Valley, AI agents are increasingly moving beyond assistance into execution. Alex Balazs, chief technology officer of Intuit, said AI tools can now close financial books on behalf of the company.
He also highlighted dramatic advances in AI driven software development.
“The days of AI assisting you with coding have become you are assisting the AI with coding,” Balazs said, signalling a reversal in traditional developer workflows.
Bret Taylor, co founder and CEO of Sierra and chairman of OpenAI, reinforced that shift. Some members of his team, he said, no longer write code directly but instead review and refine output generated by coding agents.
Security and governance hurdles
Despite the enthusiasm, executives remain cautious. Cybersecurity risks, authentication controls and governance frameworks are still evolving.
“There are issues with allowing agents to be able to do their work properly, even authentication,” Carvelli noted. “The tools need to be a certain level of security, a certain level of capability.”
Employee resistance also remains a factor. While many workers rely on AI daily, fears persist about job displacement and shifting skill requirements.
Meanwhile, financial markets are watching closely. As AI agents grow more capable, some software and data companies face investor scrutiny over whether AI platforms could compete with or replace traditional tools. Yet summit attendees suggested that the immediate priority is integration, not wholesale replacement.
A permanent shift
The consensus emerging from the summit is clear. AI agents are no longer a peripheral technology waiting for validation. They are embedded in enterprise operations across industries, from finance to automotive retail to cloud software.
The debate has moved from if to how. Companies are racing to harness productivity gains while addressing security and governance concerns.
For many executives, the transformation is already underway, and there is no turning back.
Photo Credit: DepositPhotos.com