Can employees discuss their AI use without putting themselves at risk?
An “IA on the job” programme can make tools accessible without making their use open to discussion within the team. The management question is specific: how can an employee report problematic AI use without fearing that they will appear incompetent?
What the literature establishes
Amy Edmondson (1999, Administrative Science Quarterly) identifies an association between team psychological safety and learning behaviour. Psychological safety denotes a shared belief that a team is safe for interpersonal risk-taking, including asking questions or acknowledging mistakes. Applied to AI, this research invites managers to examine the conditions under which employees can describe a difficulty, rather than simply whether training is available.
A manager’s openness must be credible
James Detert and Ethan Burris (2007, Academy of Management Journal) show that perceived managerial openness is associated with employees voicing suggestions for organisational improvement, with psychological safety playing a mediating role. An invitation to speak therefore involves more than providing a reporting channel. For “IA on the job”, this mechanism suggests paying attention to how managers respond when an employee reports an unsuccessful trial or challenges a proposed use.
The premature conclusion
One might conclude that every experiment should be encouraged and all judgement suspended. These studies justify neither an absence of rules nor indifference to the consequences of use. The useful distinction is between receiving a report in a way that enables the facts to be examined and authorising a practice, which remains subject to professional requirements. (our executive and employee training programmes)
What the evidence cannot promise
These studies do not concern generative AI, and their findings do not demonstrate that a discussion forum will reduce AI-related incidents. More reported difficulties may reflect greater willingness to speak, but also more problems: a count alone cannot distinguish between them. Conversely, an absence of reports proves neither proficiency with the tool nor an absence of use.
A practical check in Sion
In an SHR “IA on the job” workshop in Sion, teams from hydropower, healthcare, viticulture and Alpine tourism could examine fictional situations relevant to their activities, without sensitive data. For one month, each difficulty subsequently reported would receive a minimal record: the issue raised, the manager’s response and the agreed follow-up, without ranking individual employees. The measure would be the proportion of reports receiving an explicit response and a documented follow-up, verifiable in this register. It would assess whether concerns are actually addressed, not the overall safety of AI use or, on its own, employees’ freedom to speak. To go further: explore the AI on the Job training in Sion in the canton of Valais, or browse our executive and employee training programmes in Switzerland.
In pictures: AI on the Job in Sion
- ai on the job
- Sion
- research
- training sion