AI at Work: Your People Are Ready, Your Company Is Stuck

Monday, 8:40 a.m., third floor

“Look.”

Lucas, head of operations at a mid-sized industrial company, leans toward the screen his financial controller is holding out. An AI agent has just compiled the weekly report for the group’s four plants. Thirty-two consolidated tabs, variances explained, anomalies flagged in red. This work usually takes a day and a half.

“Ten minutes,” the controller says. “I built it over the weekend.”

Lucas smiles. He is genuinely impressed. Then he says the sentence I hear in almost every company I work with:

“Great. Let’s talk about it after the quarter closes. Until then, we keep the current process. I don’t want any surprises in front of the board.”

The controller closes his laptop. The following Monday, he builds the report by hand. The Monday after that, too. By the third Monday, he stops showing anything to anyone.

Lucas doesn’t exist under that name. He is a composite of dozens of executives I have met, and I bet at least one of them sits on your leadership team. Maybe even in your mirror.

What Microsoft measured, and what it left out

On May 5, 2026, Microsoft released its annual Work Trend Index. The company had Edelman Data x Intelligence survey 20,000 knowledge workers who use AI, across ten countries including France, between February and April 2026. It cross-referenced those answers with trillions of anonymized Microsoft 365 signals (Microsoft WorkLab).

The core message fits in one sentence: people are ready, the systems around them much less so. In other words, the brake on AI sits less in your employees’ heads than in your processes, your metrics and your management habits.

Before applauding, let’s put a few limits on the table.

Microsoft sells the agents whose growth it measures. The number of active agents in Microsoft 365 grew 15-fold in a year, and 18-fold in large enterprises (Microsoft WorkLab). A vendor that concludes its customers should reorganize around its products deserves a careful reading.

The sample only includes AI users. Anyone who answered “never” was screened out at the start. So you will find nothing here about the part of your workforce that ignores the tool or refuses it.

Everything is self-reported. Respondents describe their own practice, their manager, their company culture and the impact they perceive. Microsoft says so in its methodology: the relationships shown are statistical associations, with no demonstration of cause and effect (Microsoft WorkLab).

These limits don’t make the study useless. They simply require you to read it with method.

67 versus 32: reading the number without telling yourself a story

The report’s headline figure: organizational factors (culture, manager support, talent practices) account for 67% of AI’s reported impact, compared with 32% for individual factors such as mindset or behavior (Microsoft WorkLab).

Beware the quick read. That 67% doesn’t mean two-thirds of the problem comes from the organization. It is a share of relative importance in a statistical model, a random forest tested on 29 factors and nearly 20,000 responses. The most predictive factor, the organization’s AI culture, carries about two and a half times the weight of the top individual factor.

What should you take from it? When an employee says AI really helps them, their work environment predicts that perception better than their own enthusiasm does. The result tells you where to look first, without proving that changing the culture will produce the expected impact.

This reading matches something I have argued for years. The invariant of innovation has never been technology. It is the human being, with their fears, their biases and their emotions, from the intern to the CEO. An organization that buys licenses without touching its culture is buying a tool it cannot absorb (My book, chapter 4).

This book is for CEOs, managers, and innovators ready to challenge everything and create exponential growth.

Be the first to get the english version!

The Transformation Paradox, French edition

Microsoft has a name for what Lucas’s controller goes through every Monday: the Transformation Paradox. 65% of AI users fear falling behind if they don’t adapt quickly. At the same time, 45% say it feels safer to focus on current goals than to redesign work with AI. And only 13% say they are rewarded for reinventing their work even when the results don’t follow (Microsoft WorkLab).

Read the first two numbers side by side. Your teams are afraid of standing still, and even more afraid of moving. They choose the safety your metrics offer them.

In France, the top of the company looks even less aligned. 53% of French AI users fear not adapting fast enough, 43% prefer to stick to their operational goals, 12% say they are rewarded for reinvention, and only 20% consider their leadership clearly aligned on AI, compared with 26% globally (Microsoft Source).

One employee in five. The other four watch a leadership team sending mixed signals: “innovate” in the speeches, “hit your numbers” in the annual reviews. Caught between the two, they pick the numbers. In their shoes, I would do the same.

Panic about AI is directly linked to the survival instinct. When leadership doesn’t say clearly what AI will be used for, the vacuum fills with worst-case scenarios: have I become obsolete, is my job at risk? The rule of trust stays simple: I say what I do, and I do what I say. La Poste, the French postal service, learned this the hard way. Its leadership defined an AI strategy for organizing mail carriers’ routes in July 2023, and only presented it to staff as a done deal at rollout in January 2024. The immediate result: a strike (My book, chapter 14).

Author, editor, director, orchestrator: which role do you allow your teams?

The report describes several ways of working with an agent. In its French edition, Microsoft names four (Microsoft Source):

  • the author produces the work and asks AI for occasional help;
  • the editor sets the intent, lets AI produce, then reviews and validates;
  • the director writes a brief and delegates entire tasks, which run in the background;
  • the orchestrator designs a system in which several agents work in sequence and only escalate exceptions.

Microsoft itself states that this framework is illustrative: it rests on no direct measurement (Microsoft WorkLab). It remains useful for one simple reason. Moving from author to orchestrator requires a permission individuals cannot grant themselves. Lucas’s controller was ready to become a director. His organization sent him back to the author role, and even below it.

Moving up a level also demands more judgment. The users surveyed rank quality control of AI output (50%) and critical thinking (46%) as the top human skills. In France, those scores drop to 43% and 37%. And 85% of French respondents say they treat AI output as a first draft they remain responsible for (Microsoft Source).

Keep in mind the study Harvard Business School ran with BCG on 758 consultants. On tasks inside AI’s competence frontier, consultants using AI completed 12.2% more tasks, 25.1% faster. On a task deliberately chosen outside that frontier, they were 19% less likely to reach the correct solution (HBS). An orchestrator without critical thinking industrializes their mistakes.

Between the leadership team and the controller, there is Lucas. He is the one who turns a strategy into a Monday morning.

A Microsoft study of 1,800 employees, cited in the report, measures what changes when a manager uses AI in front of the team: 17 points more on perceived AI value, 22 points more on critical thinking, 30 points more on trust in agents. When that manager creates psychological safety around experimentation, employees are 1.4 times more likely to be frequent users of agents (Microsoft WorkLab).

Among “Frontier Professionals,” the most advanced users, 85% say their manager openly uses AI, compared with 64% of the others. And 26% say they are rewarded for reinventing their work regardless of outcome, compared with 11% (Microsoft WorkLab).

Nothing new under the sun. In 1999, Amy Edmondson studied 51 teams in a manufacturing company. Psychological safety, the shared belief that the team is safe for interpersonal risk-taking, predicted the team’s learning behavior, and that behavior explained part of its performance (Administrative Science Quarterly). Without that safety, nobody dares tell the CEO the project is on fire. And nobody dares show the agent they built over the weekend (My book, chapter 9).

Twenty dollars and an expense report

Hugues, an employee whose testimony I share, needed a twenty-dollar ChatGPT subscription. Purchasing told him not to expect anything for three months. With his manager’s approval, he used his company credit card. His first expense report caused a scandal in HR. The manager had to justify the decision. The leadership team experienced the episode as a wake-up call, and the purchasing procedure was changed within 24 hours (My book).

When an organization can’t remove a twenty-dollar obstacle, it won’t remove a strategic one.

The macroeconomic data points the same way. In Denmark, Anders Humlum and Emilie Vestergaard linked AI adoption surveys to administrative labor market records. Two years after ChatGPT’s launch, they find no measurable effect on earnings or hours worked, and rule out effects larger than 2% (NBER). The tool has spread. Organizations have not yet turned that usage into results.

What I would do on Monday morning in Lucas’s place

If Lucas called me, I wouldn’t ask him which tool to buy. I would suggest the five steps I use for AI adoption:

  • clarify the vision: does AI change what the company does, or only how it does it?
  • communicate that vision by answering survival questions first, after measuring the teams’ baseline sentiment;
  • hand exploration to a small team of enthusiasts, which obviously includes the financial controller;
  • deploy what has been validated, measuring before, during and after;
  • measure sentiment again, then restart the loop.

In practice, the controller’s agent would run in parallel with the current process for four weeks. We would count the time saved and the accuracy gaps. At the board meeting, Lucas would present both reports side by side. The surprise he feared would become an experiment he leads.

And I would ask him to change one metric in his teams’ annual reviews: the documented attempt, including when it fails. As long as your evaluation criteria reward only the quarter’s results, your most capable people will take their talent elsewhere.

The essentials

Remember four things:

  • The 2026 Work Trend Index links AI’s reported impact more to the work environment (67%) than to individual qualities (32%). This is a correlation drawn from a self-reported survey, and it demonstrates no causality.
  • In France, only 20% of AI users consider their leadership clearly aligned. Silence at the top feeds fear, and fear freezes teams.
  • The frontline manager is the most concrete lever: a manager who uses AI in front of the team and protects experimentation changes behavior.
  • Test your organization on a twenty-dollar obstacle. If it fails, it will fail on strategic issues too.

Your employees have already covered part of the distance. It is up to you to decide whether your organization follows them or holds them back. Innovate or agonize: the choice is yours.

References

Picture of Philippe Boulanger

Philippe Boulanger

Philippe Boulanger, international speaker on innovation and artificial intelligence, author, advisor, mentor and consultant.

Latest POSTS

41 Shades of Blue: Google’s $200 Million A/B Test

By testing 41 shades of blue, Google claims it earned an extra $200 million a year. Behind the cult A/B testing anecdote lies a debate every leader has to settle: when to trust the data, and when it stops you from innovating.

Read More »

Drone vs Camera Robot: Who Films Where the Sky Is Banned?

Banned from flying over crowds, drones run into regulation and noise. The Beni camera robot films from the ground. A textbook case of the framing bias.

Read More »

Friday the 13th: why superstition pays off for innovators

Friday the 13th destroys nothing. People who fear it change their behavior, and that change can be measured. Accidents, stock markets, the Jerusalem virus, a $900 million figure nobody verified: what superstition reveals about your innovation decisions, and how to use it.

Read More »

Naming an innovation: what Sapir showed in 1929

In 1929, Edward Sapir showed that most people link the invented word “mal” to a large table and “mil” to a small one. Almost a century later, sound symbolism still shapes how we perceive product names. What the numbers really say, what marketing took from it, and how to test a name before you pick it.

Read More »

AI-Generated Code: Who Still Understands Your Software?

A year after handing 100% of its code to AI, a company finds that nobody can explain it anymore. An anonymous testimony, the METR, GitClear and Veracode studies, and the method to stay out of the labyrinth.

Read More »

Why Coca-Cola’s Inventor Never Got Rich

John Pemberton invented Coca-Cola in 1886 and died poor two years later. Thirty-three years after the first glass, the company was worth $25 million. What this story reveals about the difference between invention and innovation, and about the Pemberton who may be sitting in your own company.

Read More »

Rare-Earth-Free Motors: France Had a 14-Year Head Start

Renault has built rare-earth-free motors since 2012. Then the stator of its next E7A motor went to China. How a technological lead gets lost in purchasing trade-offs.

Read More »

Leeuwenhoek: the cloth merchant who saw microbes first

In 1676, a Delft draper with no university education discovered microbes. The Royal Society didn’t take his word for it: it checked, then elected him. What the accurate story of Leeuwenhoek teaches leaders about ideas that come from outside.

Read More »

Are you a rule breaker?

You weren’t supposed to find this.

But here you are, because you did what most people don’t: you questioned, you explored, you clicked the thing you weren’t sure you should click.

That’s Innovational Intelligence™ in action.

Most people stay inside the lines. Follow the expected path. Click the obvious buttons. Accept things as they are.

Not you.

You’re one of those rare minds that refuses to accept “this is how it’s always been done.”

We need more people who think like you.

So here’s your reward for coloring outside the lines:

Get VIP pre-release access to the next assessment on Innovational Intelligence™:

You’ll be the first to know when it’s available.

Keep breaking rules. The world needs what you see.

Your daily dose (or almost) of Innovational Intelligence™!

Receive in your inbox my latest post before everyone else!