Managers Use AI More as Enthusiasm Cools, Wharton Survey Shows

Managers are using artificial intelligence more frequently even as enthusiasm cools, while surveyed organizations shift toward reviewing, approving and governing AI-generated work, a new survey shows.

Among managers and higher-ranking employees, 62% said they use AI daily, up from 46% the previous year. The share describing themselves as excited about AI fell eight percentage points to 51%, while 59% remained optimistic.

Eighty percent said AI had produced positive financial returns, although only 28% described those returns as strongly positive. The findings do not establish that AI caused those returns.

The survey points to a change in how departments use the technology. Seventy-two percent said their focus had shifted from creating outputs to managing AI-generated work. That figure reached 82% among vice presidents and higher, compared with 63% among managers and directors.

Prasanna “Sonny” Tambe, a Wharton professor, said, “The amazement is kind of wearing off a little bit.” He also said, “This technology is moving so fast and our organization is not ready to keep up.”

Review and accountability remain central concerns. When asked who would be responsible for a consequential error by an autonomous AI system, 28% named a governance or risk team, 26% named the deploying employee’s manager and 25% named the employee.

Forty-eight percent said AI performed work that required human approval before execution. Another 35% said AI provided suggestions or drafts while a person made the final decision, while 17% reported autonomous operation with monitoring or guardrails.

Tambe said organizations must determine “what is it that we can verify that agents can do.” Jeremy Korst, a Wharton lecturer and GBK partner, described the management challenge as deciding, “How do I prioritize what I should be doing?”

Staffing and workplace design are also changing. One-third of managers and higher-ranking employees said their departments had slowed or frozen hiring because of AI. Twenty-four percent expected fewer entry-level hires over the next two to five years, compared with 14% who expected fewer executive hires.

Managers were more likely than entry- and mid-level employees to say workers had received adequate AI training, at 78% versus 56%. The groups were not matched within the same companies.

Fifty-two percent said AI had changed their department’s management structure, while 73% said workflows had changed significantly. Departments measuring growth outcomes reported strongly positive returns more often than those measuring efficiency alone, at 35% versus 13%.

Separate Wharton research examined overreliance on AI in three experiments involving more than 1,300 participants and nearly 10,000 trials. Access to correct AI answers increased accuracy by 25 percentage points, while access to incorrect answers reduced accuracy by 15 percentage points compared with participants who had no AI access.

Gideon Nave, a Wharton associate professor, described “cognitive surrender” as adopting an AI decision without recognizing that the transfer of judgment has occurred. Because each annual wave surveyed a different group, the results reflect shifts across management ranks rather than changes among the same executives.