'Transform or die' is a sales pitch

artificial-intelligencechange-managementpsychologyconsulting

'Transform or die' is a sales pitch

Fear moves budgets, not people. What psychology says about why AI rollouts sold as transform-or-die meet resistance, and what helps teams get from threat to agency.

7-min read

Table of Contents
· 8 min · 1,510 words

‘Transform or die’ is a line I’ve heard in a lot of kick-off meetings, usually from someone who’d be billing the transformation. It works as a sales pitch because fear moves budgets. It works badly on the people who are supposed to do the transforming.

If you roll out AI like a compliance topic, with mandatory trainings, a bit of pressure and a deadline, people learn what that setup teaches them: to avoid it, to get cynical, or to resist in a polite, professional way. That’s a rational response, and psychology has been describing it for about seventy years. So before anyone writes another change plan, I think it’s worth looking at what we’re actually up against, because it’s older and deeper than AI.

What people say they’re worried about

Pew found in 2025 that more Americans are concerned than excited about AI in daily life, and their worldwide survey that October showed the same pattern in most countries.12 KPMG’s global trust study lists the worries, and they’re pretty consistent:

  • losing their job, or part of it
  • losing the human part of their work
  • outputs they can’t trust
  • their data ending up somewhere it shouldn’t
  • losing control over their work and their reputation3

That’s the list people will give you in a survey. Underneath it sit a few older mechanisms, and those are the ones a training module never touches.

Losses weigh double

Kahneman and Tversky showed in 1979 that people feel a loss roughly twice as strongly as a gain of the same size.7 Samuelson and Zeckhauser found the everyday version a few years later: given a choice, people stick with whatever’s already there, even when the alternative is better on paper.8

Put that next to an AI rollout. The gain is vague and in the future (“you’ll be more productive”). The loss is specific and right now: the way you’ve worked for ten years, the shortcuts you know, the bit of the job you’re proud of. So people aren’t weighing two equal things and choosing badly. They’re doing the sum the way every human brain does it.

Your brain reads it as danger

Social threats run on much the same wiring as physical ones. Eisenberger and Lieberman’s work on social exclusion found that being left out lights up areas that also process physical pain.9 David Rock turned that line of research into the SCARF model: five things people guard at work, and when one is under threat, the brain shifts into defence and thinking gets narrower.10 AI pokes at all five at once.

  • Status: the skill that made you the expert is now a button anyone can press.
  • Certainty: nobody can tell you what your job looks like in two years, including your manager.
  • Autonomy: the tool, the deadline and the training were chosen for you.
  • Relatedness: roles in the team shift, and it’s not clear who’s still needed.
  • Fairness: if the work gets faster, who keeps the saved time, you or the budget?

Deci and Ryan’s self-determination theory comes at it from the other side. People stay motivated when three needs are met: autonomy, competence and feeling connected to others.11 A badly run rollout takes a swing at all three in the same quarter. I wrote about the identity side of this earlier in the year, when the skills you spent years on turn into a button; that piece is When the Floor Moves.

Push, and people push back

Jack Brehm called it reactance in 1966: when people feel a freedom is being taken away, they push to get it back, often by doing the opposite of what they’re told.12 ‘Transform or die’ is close to a textbook trigger. It removes the choice and adds a threat, and then everyone’s surprised when the team finds clever reasons why the pilot can’t start yet.

The other end of the same mechanism is worse. Seligman and Maier’s early experiments on learned helplessness showed that a threat you can’t control doesn’t make you work harder; eventually you stop trying.13 In an office that looks like people who’ve gone flat. They attend the training, click through it, and change nothing. From the outside it reads as not caring. Mostly they’ve just stopped believing their effort makes a difference.

The commitment nobody says out loud

Robert Kegan and Lisa Lahey spent years studying why people who really want to change don’t. Their answer is the ‘immunity to change’: next to the goal you state, there’s a hidden commitment pulling the other way, and it’s usually protecting something.14 “I want to use AI more” sits right next to “I’m committed to being the one who knows the answer”, or “I’m committed to never looking slow in front of the team”.

You can’t train someone out of a commitment like that. You can only make it visible and give them a safe way to test whether the fear behind it holds up.

Change has an ending first

Kurt Lewin’s unfreeze, change, refreeze is the classic model: people first let go of how things were, live through a messy middle, and only later settle into something new.4 William Bridges added the part most change plans skip. A change is the external event; the transition is what people go through inside, and it starts with an ending, then a confusing neutral zone, and only then a new beginning.15 The Kübler-Ross curve, borrowed from grief, describes how people loop through denial, anger and bargaining on the way, and not in a neat order.5

Most AI programmes I’ve seen start at the new beginning. Here’s the tool, here’s the use case, go. Nobody names what’s ending, so people stay stuck there.

Resistance is information

Jeffrey and Laurie Ford argued in HBR back in 2009 that resistance is mostly feedback, and that the change agents themselves help create it through broken agreements and unclear communication.616 When a team pushes back on AI, it usually points at something nobody clarified: who’s accountable when the output is wrong, what’s allowed, what happens to the saved time, whether there’s time to learn at all. Crush it and you lose the most useful feedback you’ll get.

Why the fear pitch sells anyway

The big consultancies sell transformation as a programme: a burning platform, a maturity model, a roadmap, a change office, and a team on day rates to run all of it. The fear framing isn’t an accident. It’s what makes a programme that size feel necessary, and the research above says it’s also what makes it fail.

What I see work

What helps is mostly cheap, and it maps pretty directly onto the threats.

  • Status: make your experts the reviewers and teachers of the new tools, so their judgement becomes more valuable.
  • Certainty: say plainly what’s decided and what isn’t, with dates, even when the answer is “we don’t know yet”.
  • Autonomy: let people pick where they start and which task they try first.
  • Relatedness: learn in pairs or small groups, so nobody’s figuring it out alone in front of a screen.
  • Fairness: say out loud where the saved time goes before anyone has to ask.
  • Competence: start with something small that works in the first week, so there’s a win before there’s a mandate.

And name the ending. One honest conversation about what people are losing does more than three trainings about what they’re gaining. Moving people from threat to agency is mostly about giving them a bit of control back, and none of that needs a hundred-person programme.

If you’re leading this right now, which of these is loudest in your team: status, certainty, or the fairness question nobody wants to ask in the meeting?

Read more

  1. How Americans view AI and its impact on people and society, Pew Research Center, 17 September 2025
  2. How people around the world view AI, Pew Research Center, 15 October 2025
  3. Trust, attitudes and use of artificial intelligence, KPMG, May 2025
  4. What is Lewin’s change theory?, Prosci, 1 October 2024
  5. Kübler-Ross change curve, Elisabeth Kübler-Ross Foundation
  6. Decoding resistance to change, Jeffrey D. Ford and Laurie W. Ford, Harvard Business Review, April 2009
  7. Daniel Kahneman and Amos Tversky, “Prospect theory: an analysis of decision under risk”, Econometrica 47(2), 1979
  8. William Samuelson and Richard Zeckhauser, “Status quo bias in decision making”, Journal of Risk and Uncertainty 1, 1988
  9. Naomi Eisenberger, Matthew Lieberman and Kipling Williams, “Does rejection hurt? An fMRI study of social exclusion”, Science 302, 2003
  10. David Rock, “SCARF: a brain-based model for collaborating with and influencing others”, NeuroLeadership Journal 1, 2008
  11. Richard Ryan and Edward Deci, “Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being”, American Psychologist 55(1), 2000
  12. Jack Brehm, A Theory of Psychological Reactance, Academic Press, 1966
  13. Martin Seligman and Steven Maier, “Failure to escape traumatic shock”, Journal of Experimental Psychology 74(1), 1967
  14. Robert Kegan and Lisa Laskow Lahey, Immunity to Change, Harvard Business Press, 2009
  15. William Bridges, Managing Transitions: Making the Most of Change, 1991
  16. Jeffrey D. Ford, Laurie W. Ford and Angelo D’Amelio, “Resistance to change: the rest of the story”, Academy of Management Review 33(2), 2008

If this poked at something you're building: talk to me.

MAKE YOUR CASE.