Forty billion dollars spent on enterprise AI. Ninety-five per cent of organizations saw no measurable impact on profit.
Gallup just worked out why. And it isn’t the technology.
Their 2026 State of the Global Workplace report carries a subtitle I had to read twice: The Human Side of the AI Revolution. I’ve spent two years arguing that AI strategy succeeds or fails on the people who have to use it. It’s a genuinely odd feeling to watch the world’s largest ongoing study of the employee experience land on the same conclusion and print it on the cover.
Here’s the line that should stop every board in the country.
In organizations that have actually implemented AI, just 12% of employees strongly agree it has changed how work gets done.
Twelve per cent.
Not 12% who like it. Not 12% who’ve tried it. Twelve per cent who say the most disruptive technology of our lifetime has changed their actual work.
So where did the other 88% go?
The technology works. The return doesn’t show up.
Let me be clear about something, because I demo AI live from the stage in every keynote and I’m no skeptic: the technology works. Large language models draft contracts, write code and synthesize research faster than any human team alive. That isn’t in question.
What’s in question is why almost none of that capability is reaching the bottom line.
The $40 billion / 95% figure comes from MIT — I’ve cited it before, in my writing on the Atlassian redundancies, and it hasn’t got less damning with age. Alongside it, an NBER survey of nearly 6,000 executives worldwide found 89% see no effect on labour productivity. And here’s the tell: even OpenAI, in its own 2025 enterprise report, concluded that the real constraint is no longer model performance or tooling — it’s organizational readiness and implementation.
Read that again. The company selling the models is telling you the models aren’t the problem.
Gallup found the missing variable. It has a name badge.
So if the technology works, and the money’s been spent, and the returns have vanished somewhere between the license and the ledger — what’s the missing variable?
Gallup’s answer is refreshingly specific. It’s the manager.
In organizations investing in AI, the strongest predictor of whether employees actually adopt it — aside from technical integration — is whether their direct manager actively champions it.
And the size of that effect is startling. According to Gallup’s Q1 2026 US data, employees who strongly agree their manager actively supports their team’s use of AI are 98.7 times as likely to strongly agree AI has transformed how work gets done, and 97.4 times as likely to say it gives them more opportunities to do what they do best.
I’ll be honest — numbers that large make me want to check the methodology, and you should read them as Gallup’s comparison rather than a law of physics. But the direction is unambiguous, and it matches everything I see in the room. The manager is the multiplier. Even the most sophisticated model can’t overcome an indifferent team leader.
Now the uncomfortable part.
Fewer than one in three employees in AI-adopting US organizations say their manager actively supports their team’s use of the technology. In Germany, it’s one in five.
We’ve built the multiplier into a switch that’s mostly switched off.
And the switch is attached to an exhausted person.
Here’s where it turns from a training problem into a leadership one.
The managers we’re relying on to champion AI are, right now, the most disengaged they’ve been in years. Gallup’s global engagement figure has fallen to 20% — the first time it’s dropped two years running. Manager engagement specifically is down nine points since 2022, sliding from 27% to 22% in a single year. The “engagement premium” that used to come with stepping up into leadership has quietly disappeared.
Sit with the trap that creates.
We’re asking the single most depleted layer of the organization to be the enthusiastic front line of the biggest change in a generation. Then we’re surprised when adoption stalls.
This is two of the Three Disruptions reshaping the future of work colliding — AI and technology meeting the rise of the empowered workforce — right at the pressure point where they land: the manager who’s meant to hold it all together and has been handed a login instead of any real support.
What the 12% did differently.
The organizations getting genuine return aren’t the ones with the busiest adoption dashboards. They’re the ones that treated AI as a human capability challenge, not a software rollout.
In my Everyday AI work I talk about comprehension over compliance — the difference between people who’ve been told to use a tool and people who genuinely understand what it can and can’t do, and have redesigned their work around that difference. That’s the gap between usage and value. It doesn’t close with more licenses. It closes with capable, engaged managers.
Which means the highest-leverage AI investment on your roadmap this year probably isn’t a model or a platform.
It’s re-engaging and equipping the managers to lead the adoption you’ve already paid for.
That’s not a soft aside to the AI strategy. Gallup’s data says it is the AI strategy.
The technology was never the hard part. The humans who have to use it always were. And that’s the most encouraging finding in the whole report — because the human side is the part you can actually change.
#FutureOfWork #leadership #AI
Boards and C-suite: if you’re staring at an AI investment that hasn’t paid for itself yet, the question for your next meeting isn’t “which tools?” It’s “are our managers engaged and equipped enough to lead this?” I run executive sessions on exactly that.
Conference and event organizers: if your audience is caught between AI hype and AI reality, this is the keynote I’m most asked for right now — including the live demonstration that shifts a skeptical room from fear to strategy.
Where AI strategy meets human reality. That’s the whole game.
