AI assistants: do teams feel able to report problems?
Building an assistant without coding makes experimentation accessible, but does not ensure that difficulties reach the person in charge. The managerial question is specific: how should problems be reported when the team knows its manager supports the project?
What the literature establishes
Amy Edmondson (1999, Administrative Science Quarterly) identifies an association between team psychological safety and learning behaviour. Psychological safety refers to a shared belief that the team is safe for interpersonal risk-taking, including asking questions or acknowledging mistakes. For an AI assistant, the hypothesis to examine is therefore organisational: reporting a difficulty may feel more exposing than quietly working around it.
The mechanism of organisational silence
Elizabeth Wolfe Morrison and Frances J. Milliken (2000, Academy of Management Review) propose a theoretical model of organisational silence: certain managerial beliefs and practices can sustain the perception that raising problems is risky or pointless. This mechanism does not require an explicit ban on criticism. When an assistant is presented as an established success, a user may hesitate to report experiences that contradict that narrative.
The premature conclusion
One might conclude that opening a feedback channel or encouraging candour is enough. These papers instead invite scrutiny of what happens after a report: listening, investigating causes, responding to the person who raised the issue, or questioning their competence. An increase in reported problems may therefore reflect greater willingness to speak up rather than declining assistant performance. (our executive and employee training programmes)
What the evidence cannot establish
These publications address neither generative assistants nor no-code platforms. Edmondson's study does not establish that a reporting mechanism causally improves assistant quality, while Morrison and Milliken present a theoretical model rather than an intervention evaluation. Feedback volume alone therefore measures neither technical reliability nor people's actual freedom to speak up.
A practical check in Sion
Within SHR's programme « Créer ses assistants et agents IA sans coder », an exercise in Sion could focus on an assistant for preparing files in hydropower, providing administrative information in healthcare, handling commercial enquiries in viticulture, or welcoming visitors in Alpine tourism. During a two-week pilot, proposed here as a protocol rather than a scientific finding, keep a log of reported difficulties without sensitive data, then calculate the proportion receiving a documented response specifying the follow-up and the person responsible. Review unanswered reports with the team: this rate checks how feedback is handled, not the assistant's reliability, and an absence of reports is not evidence of success. To go further: explore the Building AI Assistants and Agents Without Coding training in Sion in the canton of Valais, or browse our executive and employee training programmes in Switzerland.
In pictures: Building AI Assistants and Agents Without Coding in Sion
- building ai assistants and agents without coding
- Sion
- research
- training sion