Gartner published something in May that should have stopped every CEO mid-sentence. Among organizations piloting or deploying autonomous AI capabilities, roughly 80% reported workforce reductions. That part is not surprising. The part that should matter: the reductions had no measurable correlation with return on investment.
Companies that cut people and companies that did not cut people reported nearly identical ROI outcomes. The layoffs created budget room. They did not create returns.
Helen Poitevin, a VP analyst at Gartner, said it plainly: “Many CEOs turn to layoffs to demonstrate quick AI returns; however, this disposition is misplaced. Workforce reductions may create budget room, but they do not create return.”
The survey covered 350 global business executives whose organizations were actively piloting or deploying AI agents, intelligent automation, or autonomous technologies. This was not a hypothetical exercise. These were companies already in motion.
The math that should bother you
If 80% of companies deploying AI are cutting jobs, and job cuts show no correlation with improved returns, then the default enterprise AI strategy is not working. The failures aren’t dramatic. The outcome just isn’t there.
The reason is structural. Removing a person from a process does not redesign the process. It creates a hole. The AI fills some of the hole. The rest of the hole stays empty, and the institutional knowledge, the judgment calls, the edge-case handling that the person used to provide quietly disappears into the gap between what the AI can do and what the workflow actually requires.
This is the operations problem dressed up as a headcount problem. Companies are looking at AI and asking “who can we replace?” when the question that produces returns is “what changes in how we work?”
Meanwhile, KPMG did the opposite
On June 9, KPMG announced it was rolling out Microsoft Agent 365 across its entire global workforce of more than 276,000 professionals. Not a pilot group. Not a Center of Excellence. Everyone.
The same announcement included Microsoft 365 Copilot deployment at the same scale. KPMG did not frame this as a cost reduction play. They framed it as a capability play. Give every person in the firm an AI collaborator and let the operational improvements emerge from how people actually use it in their daily work.
That framing matters. Think “we bought AI to cut 15% of headcount” versus “we gave every person a tool that makes their work better.” The first strategy treats AI as a substitute. The second hands people a better tool and gets out of the way.
Gartner’s data suggests the amplifier strategy is the one that correlates with returns.
The proficiency signal
A separate Gartner finding from May reinforces this. Employees who are proficient with AI across multiple use cases are twice as likely to be highly productive, 2.3 times more likely to deliver high-quality work, and 3.2 times more likely to drive effective process improvements.
Read those numbers again. The multiplier is not in the model. It is in the person using the model. A proficient employee with a mediocre AI tool outperforms a laid-off position filled by a capable AI agent every time, because the proficient employee redesigns the process around the tool. The AI agent just executes the old process with fewer hands.
Gartner is telling you, with survey data, that the highest-ROI AI investment is not the AI. What surrounds it matters more than what’s inside it.
What this means if you are making the decision right now
If your AI strategy includes headcount reduction as a primary ROI driver, the data says you are likely to get the budget room without the returns. You will report savings. Your board will see lower labor costs. And twelve months from now, someone will ask why productivity did not actually improve, and the answer will be uncomfortable.
The companies getting real returns from AI are not the ones removing people. They are the ones changing how people work. That means new workflows and operating models where humans direct AI rather than get replaced by it.
Gartner’s own prediction compounds this: by 2027, half of enterprises without a people-centric AI strategy will lose their top AI talent to competitors who prioritize workforce enablement. The companies cutting people to fund AI are also the companies most likely to lose the people who know how to make AI work.
Every company in that survey had AI. The gap came down to what they did next — redesign operations, or subtract people and call it transformation.
One of those strategies produces returns. Gartner just showed you which one does not.