Summary
AI resistance is rarely about the technology itself—it’s a signal about how change is being experienced by people. While leaders often focus on AI’s potential for efficiency and innovation, employees are navigating a different reality: uncertainty about their role, shifting expectations, and unclear measures of performance. The real concern isn’t just job loss, but loss of control, relevance, and confidence in how work will be evaluated.
Pushback often emerges when AI is introduced top-down, without involvement or context. Teams can feel that change is happening *to* them rather than *with* them, leading to disengagement or quiet non-compliance. At the same time, if tools are complex, unreliable, or fail to demonstrate immediate value, trust erodes quickly.
What leaders sometimes interpret as resistance is often a rational response to unclear goals, poor communication, or a mismatch between technology and real workflows. Employees aren’t rejecting AI—they’re questioning how it fits into their work and whether it makes them more effective or more exposed.
Ultimately, resistance is not the barrier to adoption—it’s feedback. Organizations that listen to these signals, address concerns openly, and create clarity around roles and value will move far faster than those that simply push harder.
and What It’s Really Telling You
When your staff resist AI, it isn’t technophobia — it’s information. Here’s what’s behind the pushback, especially from younger workers, and how to lead through it.
You bring in an AI tool to make everyone’s job easier. You expect relief, maybe a little gratitude. Instead you get foot-dragging, polite scepticism, the odd eye-roll — and it’s loudest from your youngest, most digitally fluent staff, the ones you assumed would lead the charge. The easy explanation is that they’re afraid of change. The easy explanation is wrong, and acting on it will cost you.
AI Resistance is information, not fear
Start by throwing out the word “Luddite.” Your younger staff aren’t frightened of technology — they live on it. They use AI to write code, plan trips, draft messages and settle arguments. When they push back on it at work, they’re not rejecting the tool. They’re objecting to something specific: how it’s being used, what it costs, or who carries the risk.
That distinction matters. If you read their resistance as ignorance, you’ll try to overcome it — with a demo, a mandate, a pep talk — and you’ll lose. What you’re actually being handed is information. The smart move is to find out what it’s telling you.
There are three things your team can see clearly that tend to get waved away in the boardroom.
They can see how it’s made
Anyone on your team who writes, designs, codes or creates knows these tools were built by training on mountains of other people’s work — much of it taken without asking. When authors sued, Meta won the case, though the judge was careful to say the ruling didn’t make the practice lawful, only that they’d argued it badly. Anthropic paid around $1.5 billion to settle a comparable claim — a line item, not a reckoning.
To a young designer or copywriter on your payroll, being told to embrace AI can feel like being asked to celebrate the thing that quietly devalued their craft and got away with it. You don’t have to agree. You do have to understand why “it’s just a tool” lands badly.
They were told to count every gram — then watched this
Your younger staff grew up being lectured about their footprint: shorter showers, recycle, fly less, every gram counts. Then they watch AI data centres draw electricity on the scale of small cities, plus a great deal of water to keep cool. After two decades of being told restraint is a moral duty, being asked to cheer on industrial-scale consumption feels like hypocrisy.
You can’t solve the energy economics of AI from your office. But you can stop pretending the objection isn’t real. Naming it honestly earns you more credibility than waving it away ever will.
Risk is structural rather than theoretical
To understand why that risk is structural rather than theoretical, you need to understand where AI drones actually come from. According to CSIS, virtually every drone on both sides of the Ukraine conflict contains components originating in Chinese factories — carbon fibre frames, rare-earth magnets for motors, lithium-ion cells, gallium-nitride chips, and the sensors that power AI targeting systems. As recently as early 2024, nearly 89% of Ukraine’s drone component imports by value came from China, and Chinese manufacturers still control roughly 80% of global drone component production.
They’re the ones most exposed
Here’s the one that’s hardest to say out loud. Entry-level and younger workers are exactly who AI threatens first. When you ask a junior to pour their know-how into a tool that might shrink their role in two years, their reluctance isn’t Luddism — it’s self-preservation. The promised upside, where it shows up at all, tends to pool at the top; the risk lands on them. They’ve done that maths.
And this isn’t a fear they’ve invented. It’s happening in the open: Meta rolled out a programme to monitor its own employees’ keystrokes and screens to train AI agents to do their jobs, and announced 8,000 job cuts in the same breath. When your staff watch the biggest firms in the world make people train their own replacements, wariness about your new tool isn’t irrational. It’s pattern recognition.
What ignoring AI resistance actually costs you
Dismiss all this as fear and here’s what you buy: quiet non-adoption, where the tool sits unused behind a brave face; workarounds that route around your shiny new process; and, worst, the slow exit of your sharpest young people, who conclude you don’t listen. You’ll also miss the genuine risks they were flagging — because the same instincts that make them uneasy often make them right. Overriding your people doesn’t make AI adoption faster. It makes it worse.
Listening is the cheapest tool you’ve got
The fix isn’t a perk or a town hall. It’s three unglamorous things: give people a real say in how AI gets used in their work, be honest about what it will and won’t do to their roles, and share the upside instead of pocketing it. Do that and something useful happens — your sceptics turn into your best early-warning system, the ones who’ll tell you when a tool is about to embarrass you in front of a client.
None of it works while you still believe the pushback is just fear. It starts the moment you treat it as information.
You can change the terms
There’s a lever bigger than any conversation, and it goes to the root of what your team objects to. Most of the unease comes from being on the wrong side of an extractive deal: your data, and your customers’, routed through a model you don’t own, run by a company whose incentives aren’t yours. You don’t have to take that deal. You can run AI you own — open models, on your own infrastructure, where the data never leaves the building. For a lot of small businesses it can work out cheaper to run, it sidesteps a real chunk of the objections, and it puts you, not a vendor, in charge of how your information is used. That’s a whole post in itself — but it’s the first thing I’d point a sceptical team toward.
Over to you
So here’s the thing to chew on. The next time someone on your team resists a new tool, don’t ask “how do I get them on board?” Ask “what do they know that I don’t?” That one swap changes the whole conversation — and it usually surfaces a risk you’d have hit anyway, later and more expensively.
What’s the pushback you’re actually getting, and from whom? The “this feels like cheating” reaction, the values objection, and the quiet fear about jobs each need a different response — tell me which one you’re seeing most, and I’ll go deeper on it.
Be braver now.
Sources
Meta copyright ruling (Kadrey v. Meta) — PBS NewsHour: https://www.pbs.org/newshour/arts/judge-tosses-authors-ai-training-copyright-lawsuit-against-meta
Anthropic ~$1.5bn settlement — NPR: https://www.npr.org/2025/09/05/nx-s1-5529404/anthropic-settlement-authors-copyright-ai
Data-centre electricity demand — Lawrence Berkeley National Laboratory / US DoE: https://newscenter.lbl.gov/2025/01/15/berkeley-lab-report-evaluates-increase-in-electricity-demand-from-data-centers/
Enterprise AI returns concentrating in a minority (”GenAI Divide”, ~95% no measurable ROI) — MIT NANDA via Fortune: https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/
Meta monitoring employees’ keystrokes/screens to train AI agents amid 8,000 layoffs (”Model Capability Initiative”) — IPE newsletter (Substack):
For more insights from our expert team go to our events and stories.
Axel Segebrecht is founder and director of Be Braver Ltd, a UK-based technology consultancy specialising in digital sovereignty, self-hosted infrastructure, and FOSS migration for European businesses.
