AI adoption: should every use be rewarded?
When introducing an AI tool, management may be tempted to reward the employees who use it most. The managerial question is precise: does a reward tied to the number of uses encourage lasting ownership, or merely the behaviour needed to obtain it?
What the literature establishes
Edward L. Deci, Richard Koestner and Richard M. Ryan (1999, Psychological Bulletin) synthesised experiments examining the effects of extrinsic rewards on intrinsic motivation. Their meta-analysis shows that tangible, expected rewards contingent on an activity can reduce voluntary engagement with it after the reward is removed, particularly for tasks that were initially interesting. This finding does not concern AI directly, but it challenges the assumption that rewarding a behaviour necessarily strengthens the desire to continue it.
The mechanism: acting by choice or to meet an expectation
Richard M. Ryan and Edward L. Deci (2000, American Psychologist) distinguish controlled and autonomous forms of motivation in their theoretical review. A person may undertake an activity without particularly enjoying it, yet remain autonomously engaged because they recognise its purpose. For a manager, explaining why AI suits a task and allowing discretion is therefore not equivalent to awarding points for every use.
The premature conclusion
It would be excessive to conclude that every reward harms adoption, or that managers should stop setting expectations. Effects depend, among other things, on the form of the reward, its conditions and how it is perceived; informative feedback on competence is not equivalent to pressure to meet a quota. The useful distinction concerns what is recognised: a sound work-related choice or simply the frequency of use. (our executive and employee training programmes)
What these findings cannot promise
These studies measure neither generative AI adoption nor the effects of a usage bonus in a Fribourg company. Voluntarily returning to an experimental task is also not the same as using a tool professionally, which depends on access, work assignments and safety requirements. A dashboard of logins alone therefore cannot distinguish autonomous adoption from a response to an incentive.
A practical check in Fribourg
As part of SHR’s programme « Conduire l’adoption de l’IA dans son équipe », an agri-food team, a bilingual industrial workshop or a service within Fribourg’s university hub could examine this distinction using an authorised task involving no sensitive data. If a temporary usage incentive already exists, compare a period when it applies with an equally long period after its announced withdrawal, while maintaining access and support. Measure, in aggregate and without individual rankings, the proportion of eligible tasks for which AI is chosen, with a brief explanation of the choice or non-use in French or German, then check a sample against the same quality criteria. A decline after withdrawal would not by itself establish a motivational effect, but it would require a distinction between use sustained by the reward and use that continues without it. To go further: explore the Leading AI Adoption in Your Team training in Fribourg in the canton of Fribourg, or browse our executive and employee training programmes in Switzerland.
In pictures: Leading AI Adoption in Your Team in Fribourg
- leading ai adoption in your team
- Fribourg
- research
- training fribourg