Leadership
AI adoption is behaviour change, not an IT project
The biggest barriers to AI are not technical but in how people think, work and trust. That is also where the biggest lever is.
In one of the largest generative AI rollouts to date, at Novo Nordisk with 20,000 employees, something emerged that I see again and again in practice: the hardest challenges are not technical. They are about behaviour, trust and leadership.
Five myths that slow adoption
IMD researchers and the leaders responsible at Novo Nordisk describe five stubborn assumptions:
- "It's all about efficiency." Employees saved 2.17 hours a week on average, but satisfaction was three times more strongly linked to better quality of work. The biggest return is often better work, not faster work.
- "Adoption will rise on its own." Usage dropped after about three months. Those who pushed through got better results later. Timed training, real use cases and champions restore momentum.
- "Digital natives will lead." Experienced employees actually gained more in productivity and quality, because they spot opportunities and judge outputs better.
- "There is no resistance in a digital culture." Some people feared being seen as lazy or unethical for using AI. Normalising use, offering safe space and clear agreements help.
- "One rollout fits all." Different teams have different needs. Tailoring by function works better.
Don't say AI won't cost jobs
Amit Joshi warns leaders against the reflex of promising that AI will not cost jobs. Employees don't believe it, and distrust slows adoption. Better is an honest, joint conversation about how roles will change. He also argues for transparency about how AI reaches its outputs, for being specific about the upside per role, and for a safe space to experiment. "Shadow AI" deserves attention too: tools people use on their own initiative, often on personal devices. Bring them into the light rather than ignoring them.
Four questions for responsible AI
Trust also requires a clear moral compass. Hoque, Sanklecha and Scade propose four questions to ask again and again at executive, programme and technical level:
- What's in it for us, also in the long term?
- What are the likely consequences for everyone affected?
- What principles does this uphold or violate?
- What does this say about who we are?
They cite IKEA, which reskilled employees whose work was changed by AI rather than letting them go. That says something about an organisation's identity.
What I advise leaders
- Start small and accessible. A few simple applications people use every day build trust faster than one big project.
- Invest in champions. A network of colleagues who go first, share what works and answer questions is the strongest engine for adoption.
- Make good examples visible. Shared prompts and successful applications save others time and normalise use.
- Measure the effect. Record a baseline up front, so success can be proven and scaling can be funded.
Sources
- Wade, M. R., Trantopoulos, K., Navas, M. & Romare, A. (2025). 5 myths holding back GenAI in the workplace. I by IMD, 19 August 2025.
- Joshi, A. M. (2026). Bosses: Stop telling staff that AI won't take their jobs. I by IMD, 8 April 2026.
- Hoque, F., Sanklecha, P. & Scade, P. (2025). Four questions to guide the implementation of responsible AI. I by IMD, 25 July 2025.
Read next
Let's talk.
About a role, an advisory question or a workshop for your team on data, analytics, AI and FP&A.