According to Fortune's coverage of the Writer and Workplace Intelligence 2026 Enterprise AI survey — which polled 2,400 knowledge workers across 30 industries — 29% of all employees admit to actively sabotaging their company's AI strategy. Among Gen Z workers specifically, that number jumps to 44%. The generation that grew up with smartphones, that has never known a world without the internet, is the cohort most actively resisting AI adoption at work. I want to talk about why — because the answer tells you everything about what goes wrong when organisations deploy AI without genuine change management behind it.
Sabotage is not the problem. It's the symptom.
When nearly half of your youngest workforce is actively working against your AI rollout, the instinct is to treat that as a discipline problem. To tighten enforcement, increase monitoring, escalate consequences. But resistance of this scale and this consistency isn't a behaviour problem — it's a communication failure that happened weeks or months before any sabotage began.
The survey tells us exactly why workers are sabotaging. 30% cite fear that AI will eliminate their job as the primary motive. Not laziness. Not technophobia. Not generational stubbornness. Fear. Specifically, fear that the tool they're being asked to champion will be used as evidence that they're no longer needed.
And here's the brutal truth: in many cases, that fear is completely rational. The same executives asking employees to adopt AI are simultaneously announcing AI-related layoffs. 60% of C-suite leaders in the survey say they plan to let go of employees who won't adopt AI. 69% say AI-related layoffs are already underway. When you tell your people to use the tool that's replacing their colleagues, and then ask why they're not enthusiastic about it — you've answered your own question.
Why Gen Z, specifically?
Gen Z's resistance makes more sense when you understand where they sit in the organisation. They're disproportionately in entry-level and junior roles — the exact positions most exposed to automation. They're early in their careers, which means they have the most to lose from a structural shift in what employers value. And they entered the workforce during a period of near-constant economic disruption: a pandemic, a cost-of-living crisis, and now an AI transition, all before their thirtieth birthday.
They're also, paradoxically, the cohort most likely to understand what AI actually does. They're not afraid of it because it's unfamiliar — they use it personally, constantly. They're afraid of it because they understand it well enough to see exactly how it could be used to restructure the roles they currently hold. That's not technophobia. That's informed self-preservation.
Coercion doesn't solve resistance — it accelerates it. The loop only breaks when the underlying fear is addressed directly.
What the sabotage actually looks like
The methods workers use to undermine AI adoption are worth understanding, because they're not all dramatic. Some are: entering proprietary data into unsanctioned AI tools, intentionally producing low-quality output from AI systems, outright refusing to use tools at all. But others are subtler — using workarounds, reverting to manual processes while technically complying with policies, or simply never flagging when an AI output is wrong.
That last one is particularly costly. An organisation that thinks its AI adoption is proceeding smoothly, because the tools are being used, may not realise that outputs are being quietly discarded or corrected downstream, that the efficiency gains aren't actually materialising, that the adoption is a performance rather than a reality. Compliance theatre doesn't show up in deployment metrics.
Layoffs are not a viable AI adoption strategy. The leaders redesigning operations with human-agent collaboration at the centre are the ones compounding their advantage. — May Habib, CEO, Writer
What good adoption communication actually looks like
I've been doing change management for over two decades. I have never seen a technology rollout succeed when the communication strategy was built on threat. Not once. What I have seen work — consistently — is this: tell people the truth early, tell them what it means for them specifically, give them a genuine role in how the change is shaped, and invest in building the capability they need to succeed in the new environment.
For AI adoption in particular, that means being explicit about what is and isn't being automated. Which tasks will change, which roles will evolve, which skills will be more valuable, not less. It means running training before the tools go live, not after. And it means giving people time to experiment, make mistakes, and build confidence in a low-stakes environment — because fluency with AI tools doesn't come from being told to use them. It comes from practising with them safely.
The organisations seeing strong adoption rates right now aren't the ones threatening their employees. They're the ones that invested in genuine preparation — the communication, the training, the change support — before they asked people to change.
Resistance is always information
44% of your Gen Z workforce pushing back on AI adoption is not a Gen Z problem. It's a signal. A very loud, very expensive signal that the people side of this change wasn't managed well enough before the technology arrived.
The good news is that signals can be acted on. It's not too late to go back and do the work that should have happened earlier — to listen, to explain, to train, to redesign. But it requires treating the resistance as data, not defiance. And it requires leaders who are willing to ask the harder question: not "why won't they adopt?" but "what did we tell them that made this feel threatening?"
