PHILADELPHIA, PA — Only 28% of enterprise leaders report strongly positive returns on artificial intelligence investments despite widespread adoption, according to a new University of Pennsylvania Wharton School study that links better financial outcomes to redesigned workflows, business growth metrics, and centralized AI oversight.
The fourth annual study by Wharton Human-AI Research (WHAIR), conducted with market research firm GBK Collective, found that 80% of enterprise leaders reported some positive return on AI investments. Yet just 7% of organizations met all three criteria researchers identified for successful adoption: demonstrating business value, transforming workflows and aligning operations around the technology.
The findings, based on a survey of 1,416 U.S. enterprise leaders and entry- and mid-level employees, suggest that integrating AI into existing business activities has not consistently translated into substantial financial returns.
Companies measuring AI performance against growth-related outcomes were considerably more likely to report strong returns than those focused exclusively on efficiency.
Among departments tracking at least one growth measure, including revenue, sales conversion, output volume or customer satisfaction, 35% reported strongly positive AI returns. That compared with 13% among departments evaluating AI solely through efficiency measures.
Despite the difference, time savings remained the most common performance metric, used by 40% of departments. Net-new revenue was measured by 27%.
“Companies believe AI is going to transform their industries, but many are still asking it to prove itself by saving time,” said Prasanna “Sonny” Tambe, a Wharton professor and faculty co-director of WHAIR.
“The companies seeing stronger returns are going further. They are redesigning workflows, putting clear ownership and governance in place, and measuring AI against outcomes that matter to growth,” Tambe said.
Organizational structure also showed a substantial association with reported returns.
Among managers and senior executives, 58% at organizations with a newly designated chief AI officer reported strongly positive returns, compared with 24% where responsibility had been added to an existing executive’s duties. The figure fell to 8% at organizations without such a position or plans to establish one.
Companies that standardized on a single approved AI platform reported strongly positive returns at a rate of 50%, compared with 9% among organizations still experimenting without a defined approach.
The findings establish associations between management practices and reported returns, rather than proving that particular organizational structures directly caused better financial performance.
AI agents are becoming common across businesses, although their use remains heavily supervised.
Sixty-seven percent of enterprise leaders reported using AI agents to complete tasks or manage workflows. Only 28% said AI was fully embedded in core business processes, while 17% allowed agents to operate without human approval at each step.
One in five organizations met all three criteria researchers used to define workflow transformation: substantially changed processes, AI integration into core operations and use of AI agents.
Among departments meeting those criteria, 67% reported strongly positive returns, more than twice the 28% rate among enterprise leaders overall.
The study also found that companies reporting stronger AI returns were no more likely than other organizations to report workforce reductions or hiring freezes.
Instead, they were more likely to have redesigned workflows, modified management structures, monitored AI usage, incorporated AI into performance evaluations and provided employee training.
The distinction comes as companies increasingly expect AI to reshape their industries. Eighty-three percent of enterprise leaders anticipate major industry changes within two to five years.
Nevertheless, 52% said AI’s primary role over the next 12 to 24 months would be improving productivity or reducing costs, while 22% prioritized revenue growth, new products or better customer experiences.
The research also identified a significant disconnect between management’s assessment of AI implementation and employees’ experiences.
Seventy-three percent of enterprise leaders said AI had substantially changed departmental workflows, compared with 58% of entry- and mid-level employees.
The disparity was larger for training. While 78% of leaders believed employees received adequate AI instruction, only 56% of employees agreed.
Similarly, 79% of leaders believed employees were confident using AI in daily work, compared with 62% of employees. Leaders were also more likely to believe their organizations had communicated a clear long-term AI strategy, 79% versus 64%.
Employees nevertheless reported substantial benefits from the technology. Seventy-three percent said AI had positively affected their jobs, while 84% of employees using AI said it saved them time.
Workforce restructuring remains another consequence of adoption.
Thirty-seven percent of enterprise leaders reported AI-related staff reductions, while 33% reported slowed or frozen hiring.
Entry-level positions face greater anticipated pressure than senior management roles. Twenty-four percent of leaders expected fewer entry-level hires over the next two to five years, compared with 14% expecting reductions in executive positions.
At the same time, 82% said AI helped newer employees become productive more quickly, indicating that the technology may simultaneously improve early-career performance and reduce demand for some entry-level positions.
Concerns about declining skills were also widespread. Forty-eight percent of vice presidents and more senior executives worried that their own skills or cognitive abilities were deteriorating because of AI, compared with 46% of entry- and mid-level employees and 32% of managers and directors.
Among senior executives, 82% said their responsibilities had shifted toward managing AI-generated output, while 67% said they would be substantially less productive without the technology.
AI usage continued to increase even as attitudes became more cautious.
Daily use among managers and senior executives rose approximately 16 percentage points to 62%, while 91% reported using AI more than once a week.
However, the share describing themselves as excited about AI declined eight percentage points from the previous year. Cautious sentiment increased seven points, while skepticism rose eight points.
Researchers said it was the first year of the study in which AI usage increased while enthusiasm declined.
“Over four years, we have watched AI move from experimentation into everyday business operations,” said Stefano Puntoni, a Wharton marketing professor and faculty co-director of WHAIR. “What separates the leaders is whether they are changing the business around AI enough to generate meaningful value.”
The report, The Adaptive Advantage: Who’s Winning with AI and How, identifies organizational alignment and measurable business outcomes as distinguishing characteristics of companies reporting the strongest AI returns.
Support the local news that supports Chester County. MyChesCo delivers reliable, fact-based reporting and essential community resources—free for everyone. If you value that, click here to become a patron today.