AI adoption: who takes responsibility for handoffs?
A team can use an AI tool regularly while leaving the allocation of subsequent work unresolved. The managerial question is specific: when a step is automated, who becomes responsible for the transition to the next one?
What the literature establishes
Raja Parasuraman, Thomas B. Sheridan and Christopher D. Wickens (2000, IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans) distinguish several functions that can be automated: information acquisition, analysis, decision selection and action implementation. Their model shows why describing work as “automated” without specifying the function concerned is insufficient. For managers, this distinction suggests describing what the tool handles and what still needs to be assigned.
The mechanism: connecting contributions
Gerardo A. Okhuysen and Beth A. Bechky (2009, Academy of Management Annals) present a review in which coordination rests on three integrating conditions: accountability, predictability and common understanding. Routines, plans and meetings can contribute to these conditions without being universal solutions. Applied to AI, this framework suggests examining whether everyone knows what contribution to expect, from whom and when.
The premature conclusion
One might conclude that assigning someone responsibility for every AI output is enough. Yet assigning responsibility specifies neither the conditions for a handoff nor what the recipient should do next. The issue is therefore not simply to name a tool owner, but to agree on how work passes between contributors. (our executive and employee training programmes)
What these studies do not demonstrate
These two articles provide, respectively, a model of automation and a review of coordination; they are not trials of contemporary generative AI adoption. They do not establish that clarifying handoffs necessarily improves performance. Conversely, a usage dashboard alone does not reveal whether work flows smoothly between people.
A practical check in Berne
Within SHR’s “Leading AI adoption in your team” programme, one proposed exercise is to select an authorised workflow: preparing a briefing in the federal administration, synthesising documents in public health, handling a telecoms ticket or drafting a deviation report in precision manufacturing. For two weeks, the team logs every handoff following an AI-assisted step, recording its recipient and the next expected action without copying sensitive data. It calculates the proportion of handoffs for which both elements were explicit at the time of transfer, keeping the numerator and denominator auditable in the log. The result is a local coordination diagnostic, not evidence of productivity or a comparison between these sectors in Berne. To go further: explore the Leading AI Adoption in Your Team training in Bern, or browse our executive and employee training programmes in Switzerland.
In pictures: Leading AI Adoption in Your Team in Bern
- leading ai adoption in your team
- Bern
- research
- training bern