Despite pouring millions into artificial intelligence, most U.S. companies have yet to unlock the full potential of agentic AI due to a mix of technical, organizational, and regulatory barriers, according to a new survey from EY, also known as Ernst & Young, a global professional services and accounting firm.
The new data reveals a growing rift between AI ambition and execution. While 21% of organizations have already invested $10 million or more in AI — up from 16% last year — and 35% plan to do so next year, only 14% of senior leaders report that agentic AI is fully implemented in their organizations.
“AI agents can revolutionize how we work and unlock possibilities that were once unimaginable,” said Dan Diasio, EY Global consulting AI leader. “But business executives are still grappling with how to translate that potential into meaningful impact.”
The survey of 500 U.S. senior decision-makers (SVP and above) found that 97% of companies investing in AI are seeing positive ROI — particularly those that dedicate at least 5% of their total budget to AI. These high-investment firms report stronger gains in tech upgrades, customer satisfaction, and cybersecurity than those spending less.
How executives are tackling agentic AI adoption
Despite this strong ROI, just 34% of organizations have started implementing agentic AI.
Agentic AI refers to systems capable of acting autonomously to manage tasks or make decisions. The technology is being used primarily to assist internal processes (86%), including:
- Enhancing customer support (55%)
- Improving IT efficiency (55%)
- Bolstering cybersecurity (51%)
Yet adoption remains tentative. Over half of senior leaders believe their peers — and even their own colleagues — don’t fully understand what agentic AI is or how it can benefit the business.
Although 73% of executives believe that agentic AI could eventually manage entire business units, real-world concerns are slowing progress. An overwhelming 87% report barriers to adoption, including:
- Cybersecurity risks (35%)
- Data privacy concerns (30%)
- Lack of regulatory clarity (21%)
- Lack of internal policies (21%)
While fears of job loss persist — 64% believe that fear of replacement could stifle adoption — only 24% cite employee resistance as a top barrier. Still, companies are using an initiative-taking approach. 64% of leaders say their organizations will increase AI training over the next year, up from 49% a year ago.
In response to market maturation and rising demand for differentiation, organizations are increasingly building custom AI solutions in-house (64%, up from 56% last year). At the same time, fewer firms are pursuing acquisitions of AI companies (45%, down from 51%).
“While the allure of autonomous systems is strong, our findings underscore the importance of a deliberate and human-centric approach to agentic AI,” said Whitt Butler, EY Americas vice chair, consulting. “The future of work will be shaped by how well organizations prepare their people, embed responsible governance, and align AI capabilities with real business outcomes.”
Sign up
Sign up for a complimentary subscription to Digital Commerce 360 B2B News. It covers technology and business trends in the growing B2B ecommerce industry. Contact Mark Brohan, senior vice president of B2B and Market Research, at [email protected]. Follow him on Twitter @markbrohan. Follow us on LinkedIn, X (formerly Twitter), Facebook and YouTube.