The companies pouring money into AI right now are learning a hard lesson. The tool was never the hard part.
When researchers ask the people on the ground how it's going, the story changes fast. Managers see hallucinations, broken workflows, and one more login to remember. Executives see a strategic weapon. Both groups are looking at the same tool and describing two different companies.
That’s expensive. It shows up as stalled adoption, wasted licenses, and a leadership team that keeps asking why the numbers haven't moved. The missing step is diagnosis. Before you spend another dollar on software, you need a clear picture of where your team stands today, what they're capable of, and what's stopping them.
Here’s what the latest research says about this disconnect, and why fixing it has to come before your next AI purchase.
The Executive and Manager Disconnect Is Draining Your AI Budget
Every dollar spent on AI software assumes that the people using it believe it works. That assumption is falling apart inside companies right now, and the research proves it. A study from the Wharton School and GBK Collective found that 45% of executives report significantly positive returns from their AI investments. Among middle managers, that number drops to 27%.
That difference is exactly what turns a promising AI purchase into shelfware. Executives approve the budget based on strategic potential. They see the technology handle synthesis, drafting, and high-level decision support well, so their confidence climbs.
Managers are the ones who inherit the mess underneath that confidence. They're the ones stitching AI output into workflows built over years, managing teams with uneven comfort levels, and cleaning up when the tool gets something wrong in front of a client. When the tool fails, only one of those two groups has to fix it, and it's rarely the executive who approved the check.
Buying another license without addressing this divide only widens it. Your managers will keep quietly under-using the tool while your executives keep wondering why adoption looks so soft on paper. Frustration builds on both sides, and the software gets blamed for a problem it never caused.
This is precisely why diagnosis has to come before purchase. You can’t fix a problem you haven't measured, just like you can’t expect a manager who's skeptical of the technology to embrace a mandate handed down without context. Understanding where your leadership team and your frontline managers stand, not where you assume they stand, is the difference between an AI investment that pays off and one that quietly stalls for another year.
Nobody Knows Who Owns AI, and That's Why It Stalls
Only 34% of C-suite executives find it consistently clear which executive or team makes the calls on AI. That number was the lowest of every group surveyed, including board members and senior managers below the C-level.
The people closest to approving AI spending are the least confident that anyone is steering it. Board members felt clearer at 53%. Senior managers and professionals below the C-suite felt clearer still, at 57%. The farther a leader sits from daily implementation, the more comfortable they feel with the ambiguity, and the more money they're willing to authorize into that ambiguity.
This should alarm any business owner who's about to sign another AI contract. A tool purchased without a clear owner has no one accountable for whether employees adopt it, whether training happens, or whether the rollout gets measured at all.
Ownership confusion at the top doesn't stay at the top. It cascades into every department that never gets clear direction, every manager who's left guessing whether adoption is even a priority, and every employee who quietly decides the tool isn't worth the friction.
Before another purchase order gets signed, someone needs to be able to answer a simple question: Who owns this, and how will we know if it worked? If your organization can't answer that today, a new subscription won't answer it tomorrow.
Your Managers Are Quietly Rejecting the Tools You Already Bought
A survey from Nitro found that while 85% of C-suite executives report deploying AI across their organizations, only 54% of managers say AI is even a top priority. Barely half say the technology has reached at least some of their own workflows.
That number should stop any business owner mid-scroll. You can announce a rollout, celebrate the launch internally, and still have half your management layer quietly treating the initiative as optional. Among the managers whose teams have adopted AI at all, 37% primarily rely on standalone tools such as copying and pasting content into ChatGPT rather than the platforms leadership selected and paid for. Only 12% report AI actually embedded into their daily software.
The reasons behind that rejection matter as much as the rejection itself. Managers cite security and trust as their biggest barriers to adoption, followed by integration complexity and implementation costs. More than half admit sensitive documents are being uploaded to public AI tools inside their own companies, while only 43% report having clear, actively enforced AI policies. Executives, meanwhile, keep telling themselves the rollout succeeded because the contract got signed and the training session got scheduled.
This distance between announced adoption and actual behavior is precisely what a proper diagnosis catches before it costs you another renewal. Buying a second tool to fix the first tool's low adoption only adds another subscription to the pile of software your managers have already learned to route around.
What your organization needs first is an honest look at why the current tool got rejected, whether policy and security concerns are driving people back to public chatbots, and whether your managers even see AI as a priority worth their limited time. Skip that diagnosis, and your next purchase will follow the exact same path into the drawer.
90% of Executives Admit AI Hasn't Paid Off Yet
Layoffs are happening at companies that haven't seen a single dollar of AI return yet. Fortune's reporting found that a striking share of leaders privately admit their AI investments have not delivered the productivity gains they promised the board. Yet many of those same companies started trimming headcount before the technology proved itself, treating job cuts as a shortcut to the returns AI hasn't produced on its own.
The math behind that decision rarely works out the way executives hope. Researchers examined stock market reactions to AI-linked layoff announcements and found the average return sat close to zero. A handful of companies, including the financial technology platform Block, saw their stock climb on news of AI-driven staff reductions. For more than half of the events studied, though, the market reaction was flat or negative. Cutting people to manufacture a return on an unproven investment is a bet that rarely pays off, and the data backs that up.
This pattern should concern any business owner tempted to treat layoffs or a fresh software purchase as a fix for stalled AI results. The instinct to act, whether that means buying another tool or cutting a few positions to free up budget, comes from the same place: pressure to show the investment is working.
But neither move addresses the actual reason results haven't shown up. If your team hasn't been trained properly, managers don't trust the tools, and nobody owns the rollout, a new subscription or a smaller headcount won't change the outcome. It will only mean fewer people left to clean up the same unsolved problem.
The businesses that break this cycle are the ones willing to diagnose before they spend another dollar or eliminate another role. Our AI Impact Report gives you exactly that starting point. It measures where your team's skills, habits, and confidence actually stand today, so every future AI decision, whether it's a new tool, training program, or policy change, is built on evidence instead of assumption.