According to Forbes (27 July 2026), up to 70% of the value of successful AI transformations comes from people rather than technology. Cognitive neuroscientist Gleb Tsipursky, who relays this finding from BCG, distinguishes seven profiles of reaction to AI in the workplace, from the most reluctant to the most enthusiastic.
The facts
- BCG found that, in successful AI transformations, 70% of the value created comes from people-related actions rather than technology, an observation cited by Tsipursky, author of “The Psychology of AI Adoption at Work”.
- Tsipursky distinguishes seven employee profiles in the face of AI, from the alarmist, the most reluctant, driven by fear for their job, to the evangelist, the most enthusiastic, who actively encourages colleagues to adopt it.
- Roles with a heavy procedural load, accounting, legal, specialised support, express the most reservations; targeted training is often enough to answer their objections, according to Tsipursky.
Why is a single training course not enough to get AI adopted?
Each profile reacts to AI for different reasons, which makes a one-size-fits-all training course ineffective. The alarmist fears the disappearance of their job or being held responsible for the tool's mistakes, a reflex Tsipursky links to status quo bias and loss aversion. The reluctant adopter has no fear of AI itself: they fear being seen as outdated if they do not use it. In both cases, the central issue remains the same according to Tsipursky: employees want transparency about the effect of AI on their job, and the assurance that productivity gains do not automatically mean job cuts.
The approach Tsipursky recommends is to identify the most enthusiastic employees, often technical or creative profiles, and to entrust them with one or two visible, low-risk use cases. The results obtained on the ground then build the trust of the more reluctant profiles, without any need to win them over through speeches.
AIxH's view
That is exactly the logic of our AI training and upskilling programme for teams, delivered in Luxembourg and France: training all employees while taking their own reservations into account, rather than applying a single programme across the whole organisation. Our sessions first identify the readiest profiles, to launch a first visible, low-risk use case. If you want to map how the profiles in your organisation relate to AI, an initial conversation is enough to establish a diagnosis.