IA on the job: standardising without removing autonomy?
Introducing AI often brings prompt templates, shared formats and recommended workflows. The managerial question is specific: what should be standardised to safeguard work without turning assistance into a prescribed method for every task?
What the literature establishes
J. Richard Hackman and Greg R. Oldham (1976, Organizational Behavior and Human Performance) identify autonomy as one of the job characteristics that can support internal motivation through experienced responsibility for outcomes. Their model therefore does not reduce a job to its task volume or technical efficiency. Applied to AI, this perspective directs attention not only to what the tool enables people to produce, but also to what they can still decide.
A requirement can be understood without being chosen
Marylène Gagné and Edward L. Deci (2005, Journal of Organizational Behavior) distinguish autonomous from controlled motivation and explain how work requirements can be internalised to different degrees. Understanding and accepting the rationale for a rule is not equivalent to acting solely under the pressure of monitoring. In AI use, explaining the purpose of a mandatory format and acknowledging the difficulties people encounter therefore matters as much as issuing instructions.
The premature conclusion
One might conclude that everyone should freely choose their tools and methods; these studies do not warrant that conclusion. Autonomy means neither the absence of rules nor permission to bypass confidentiality, validation or safety requirements. The task is to distinguish necessary constraints from prescribed methods whose justification has yet to be established. (our executive and employee training programmes)
What the evidence cannot establish
Neither article examines generative assistants or establishes the causal effect of an AI deployment on autonomy at work. The first tests a model of job design; the second offers a theoretical synthesis of motivation. Neither tool usage rates nor the presence of an “edit” option is sufficient to demonstrate that employees have a genuine choice.
A practical check in Bâle
In Basel, the “IA on the job” programme can provide a framework for examining decision-making latitude in pharma, life sciences, Rhine logistics or an international R&D team. Across a set of assignments defined in advance, measure the proportion for which employees could choose another authorised method instead of the recommended AI workflow, without requesting an exemption. Verify each case against the written instructions, the employee’s account and the manager’s confirmation, without changing mandatory controls. This ratio does not measure motivation: it makes the gap between promised autonomy and actual permitted choice verifiable. To go further: explore the AI on the Job training in Basel, or browse our executive and employee training programmes in Switzerland.
In pictures: AI on the Job in Basel
- ai on the job
- Basel
- research
- training basel