Adopting AI: will the team still be able to work without it?
An AI tool can become indispensable before an organisation has decided how to work when it is unavailable. The management question is specific: which skills must be maintained so that the team can continue an essential activity without the tool?
What the literature establishes
Lisanne Bainbridge (1983, Automatica) identifies a classic difficulty with automation: the more a system handles routine operations, the fewer opportunities people have to practise the skills required to take over. Yet the operator is still expected to intervene when a situation arises that the system cannot handle properly. This analysis does not measure the effects of today's generative AI; it exposes a contradiction in work design.
Dependence is also shaped by the organisation
Raja Parasuraman and Victor Riley (1997, Human Factors) distinguish appropriate use of automation, excessive reliance, disuse, and deployment decisions that neglect human consequences. Their review shows that these behaviours cannot be reduced to individual preferences for technology. System reliability, workload and conditions of use help shape people's relationship with the tool.
The premature conclusion
One might conclude that everything should still be done manually to preserve skills. These papers do not justify such a rule: maintaining every action unchanged would cancel out some of automation's value. Instead, the management task is to identify activities whose interruption would be difficult to accept and the skills needed to resume them. (our executive and employee training programmes)
What the evidence cannot determine
These publications address automation broadly and predate contemporary generative assistants. They establish neither a universal frequency for practice without AI nor a general timescale for skill loss. Completing a task quickly with assistance therefore does not, by itself, demonstrate that a team could continue working with reduced system support.
A practical check in Fribourg
Within SHR's programme « Conduire l’adoption de l’IA dans son équipe », one possible application in Fribourg would be a fallback exercise without AI, developed with a bilingual food-processing or industrial company and, where relevant, a partner from the local university ecosystem. Using a fictional case outside live operations, the team would prepare an operational handover in French and German, with authorised reference documents but without a generative assistant. The manager would measure the time needed to produce a usable version and the proportion of required information correctly conveyed in each language, using a checklist and thresholds defined before the exercise. This test would not prove that AI had caused skill loss; it would reveal the team's current continuity capability and the areas requiring practice. To go further: explore the Leading AI Adoption in Your Team training in 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