Enterprise AI adoption is moving beyond experimentation, but the organizations seeing the strongest returns are doing more than adding new tools. They are redesigning workflows, establishing clear ownership, and aligning employees around how AI should be used.
That was the central message of the latest AI Horizons webinar, hosted by Stefano Puntoni and Prasanna “Sonny” Tambe, faculty co-directors of Wharton Human-AI Research, with Jeremy Korst, WG’03, founder of Mindspan Labs and partner at GBK Collective. The discussion focused on findings from their fourth annual AI adoption report — The Adaptive Advantage: Who’s Winning With AI and How.
The 2026 study surveyed approximately 1,400 employees at large U.S. companies. For the first time, it included both managers and senior leaders as well as entry- and mid-level employees, allowing the researchers to compare how different parts of an organization experience AI adoption.
Adoption Is No Longer the Main Question
AI use has continued to rise. Sixty-two percent of managers and senior leaders report using AI daily, while 91% use it at least weekly. Seventy-three percent say their department’s workflows have changed significantly because of AI, and 67% report using AI agents in some form.
Adoption alone, however, does not distinguish successful organizations from unsuccessful ones. Only 28% of managers and senior leaders say AI is fully embedded in their core business processes, and just 20% of departments meet the report’s definition of successful workflow transformation.
“Agents are quite different” from conversational tools, Tambe said during the webinar. While an assistant can often be added to an existing job without changing the underlying process, agents can carry out tasks across functions and operate with varying degrees of autonomy. That makes workflow redesign a central management challenge.
The report identifies three markers of organizational success:
- Strongly positive return on investment
- Significant workflow transformation
- Organizational alignment
Only 7% of organizations meet all three criteria. The report calls these organizations “adaptive”— companies that are not simply using AI, but changing how work is organized around it.
Growth Matters More Than Efficiency Alone
Most companies still evaluate AI primarily through efficiency measures such as hours saved, operating costs, and turnaround time. Only 22% identify growth, including new revenue, improved sales conversion, increased output, or customer satisfaction, as a primary focus.
Yet organizations that track growth are more likely to report strong returns. Only 13% of departments that measure efficiency alone report strongly positive ROI, compared with 35% of departments that track at least one growth-related measure.
Puntoni emphasized that the finding is correlational rather than causal. Still, he said it supports a broader argument: the long-term promise of AI is not limited to doing existing work faster or more cheaply. It also includes creating new products, improving customer experiences, and developing new sources of revenue.
The researchers also cautioned that productivity gains do not automatically become innovation. Employees may save time with AI, but that time often disappears into other tasks rather than being redirected toward experimentation or growth.
Ownership and Accountability Remain Unsettled
Organizations with a dedicated chief AI officer or equivalent report stronger returns than those without a designated AI leader. Fifty-eight percent of organizations with a newly created CAIO role report strongly positive ROI, compared with 24% where AI responsibilities have been added to an existing executive’s role and 8% where no such role exists or is planned.
The speakers stressed that a CAIO should not become the only person responsible for AI. Instead, the role should coordinate a broader transformation across technology, operations, finance, marketing, human resources, and other functions.
Accountability becomes more complicated as agents take on consequential tasks. No single group is identified as responsible for agent errors by more than 30% of respondents. Governance or risk teams are named most often, followed by the deploying employee’s manager and the employee. Eleven percent say no one is clearly accountable or that they do not know who is.
Leaders and Employees See Different Organizations
The report’s most consistent finding is a gap between managers and employees.
Thirty-nine percent of managers and senior leaders describe their organizations as aligned around AI, compared with 26% of entry- and mid-level employees. The gap appears across measures including confidence using AI, leadership’s long-term vision, employee empowerment, responsible AI investment, and training.
Training shows one of the largest differences: 78% of managers and senior leaders say employees receive adequate AI training, while only 56% of entry- and mid-level employees agree.
The researchers also found that sentiment is becoming more complicated. AI usage is rising, but enthusiasm is cooling. Among managers and senior leaders, the share describing themselves as excited fell eight percentage points, while cautious and skeptical responses increased.
For organizations implementing AI at scale, the implication is clear: employees need more than access to tools. They need a coherent explanation of how work is changing, what they are expected to do differently, and how the organization will support them through the transition.
The report’s central message is that AI adoption has entered a new phase. The organizations best positioned to benefit will be those that redesign core workflows, measure growth alongside efficiency, establish clear ownership, give agents meaningful but carefully governed authority, and bring employees along.
