A major new report from Globalization Partners (G-P) has put a number on something many business leaders already feel but haven’t said out loud: three-quarters of executives are disappointed with what AI has actually delivered.
G-P’s third annual AI at Work report, released May 12 and based on responses from 2,850 senior leaders across six global markets, found that 73% of executives say AI ROI is falling short of expectations. That’s a striking figure for a technology that every company on earth has been pouring money into for the past two years.
The Budget Threat Is Real
The numbers get more serious when you look at what executives plan to do about it. Nearly 70% of respondents said they are prepared to scale back AI budgets if goals are not met this year. After two years of “investing to learn,” the tolerance for slow returns is running out.
What makes this especially telling is the near-universal adoption backdrop: 100% of executives surveyed reported their organisations are already using AI. This is not a story about laggards who haven’t tried. This is a story about organisations that have invested, deployed, and waited — and are still not seeing what they expected.
The Hidden Tax Nobody Talks About
One of the more pointed findings in the report: 69% of executives say humans in their organisations are now spending more time monitoring AI outputs than they used to spend simply doing the work themselves.
That is a significant problem. When AI is supposed to create time, but instead creates a monitoring burden, the efficiency math collapses. This is what happens when AI tools are adopted without a clear workflow redesign or a strategy for where human judgement is actually needed.
The report also flags a related issue that deserves more attention than it typically gets: 88% of executives are concerned that employees are using AI to “perform productivity” without generating real business value. Nearly half — 47% — said they are very or extremely concerned this is already happening in their organisation right now.
That phrase, “performing productivity,” is a precise description of a real failure mode. An employee generates a five-page summary with AI, attaches it to an email, and the work looks impressive. But if nobody reads it, nothing changes, and no decision was improved, then the AI didn’t help the business. It just helped someone look busy.
The Shift in Ambition
Perhaps the most revealing data point in the report is a year-over-year decline in AI aggression. The percentage of respondents who describe themselves as using AI “aggressively” to innovate dropped from 60% last year to 42% this year.
That is not a sign of maturity. That is a sign of a cooling off after disappointing returns. Companies that were charging ahead are pulling back, and the ones still moving forward are doing so more cautiously.
There is also a striking finding about workforce perception: 82% of executives said AI has lowered the value placed on human workers inside their organisations. That is a people problem that will compound into a performance problem if it is not addressed directly.
What This Means for Business
The G-P report is effectively describing the gap between AI adoption and AI value. Most companies have crossed the adoption threshold. Very few have crossed the value threshold.
The difference between those two points is not about which AI tool you chose. It is about three things:
First, strategy before deployment. Most AI rollouts start with a tool and work backwards to a use case. The organisations getting real returns start with a problem worth solving and then evaluate whether AI is the right solution and which form of AI fits.
Second, workflow integration, not workflow addition. If AI gets bolted onto an existing process without changing that process, you get the monitoring burden described in the report. The process itself needs to be redesigned around what AI can reliably do and what humans still need to own.
Third, clear measures of actual output. “Productivity” is too vague. What decision got better? What customer interaction improved? What cost was reduced? Organisations that define success in concrete terms before deploying AI are the ones that can honestly assess whether they’re getting returns.
Enterprise DNA works directly with business leaders on exactly these three problems — through our advisory services for organisations designing AI strategies from the ground up, and through our learning platform for the workforce that needs to build the skills to make AI work in practice, not just in theory.
The G-P findings are uncomfortable, but they are also an opportunity. The organisations that address the ROI gap now — by building strategy, redesigning workflows, and measuring real outcomes — are the ones that will be ahead when others are still asking why their AI spending isn’t paying off.
That window is open. It will not stay open indefinitely.