AI adoption: does the way a task is presented change its acceptance?
Presenting AI as decision support does not establish which parts of the work employees consider appropriate to delegate. The managerial question is specific: for an identical task, does emphasising measurable criteria rather than professional judgment change people’s willingness to consult the tool?
What the literature establishes
Noah Castelo, Maarten W. Bos and Donald R. Lehmann (2019, Journal of Marketing Research) show that algorithm aversion depends in part on whether a task is perceived as subjective or objective. In their experiments, participants are less willing to rely on an algorithm when the task seems subjective. The authors also show that making a task appear more objective can increase this willingness: acceptance therefore depends on more than the tool’s characteristics.
A preference that can also favour the algorithm
Jennifer M. Logg, Julia A. Minson and Don A. Moore (2019, Organizational Behavior and Human Decision Processes) find that, in estimation tasks, participants can give greater weight to advice presented as algorithmic than to advice presented as human. This finding does not describe a universal preference for machines, but it challenges the idea of systematic rejection. The stated source of advice can therefore affect its influence, independently of its content alone.
The premature conclusion
One might conclude that presenting all work as objective analysis is enough to facilitate adoption. That would confuse a change in perception with an improvement in decision quality. Selecting projects or prioritising requests involves trade-offs that do not disappear when translated into numerical criteria. (our executive and employee training programmes)
What these experiments cannot establish
These studies examine reliance on algorithms and the weighting of advice, not the sustained integration of generative AI into teams in Lausanne. They establish neither that framing improves work quality nor that a stated preference becomes regular use. An adoption assessment should therefore distinguish between consulting the tool, incorporating its proposal and the quality of the outcome.
A practical check in Lausanne
As part of the SHR programme on leading AI adoption within a team, a Lausanne-based team in medtech, higher education, international sport or a Lake Geneva region scale-up could test the same fictional prioritisation case, using no sensitive data. Randomly assign participants to two accurate presentations of the task—one emphasising measurable criteria, the other professional judgment—while keeping the tool, information and opportunities for consultation identical. Measure the proportion in each group who actually consult AI before deciding, then have the decisions assessed against a common rubric by someone unaware of which presentation participants received. A difference in consultation without an improvement in quality would call for reviewing the accompanying messaging, rather than concluding that the team is adopting AI more successfully. To go further: explore the Leading AI Adoption in Your Team training in Lausanne in the canton of Vaud, or browse our executive and employee training programmes in Switzerland.
In pictures: Leading AI Adoption in Your Team in Lausanne
- leading ai adoption in your team
- Lausanne
- research
- training lausanne