Translation into curriculum development
The cross-case synthesis makes it clear that the impact of GenAI on professional autonomy cannot be reduced to the question of which tasks AI does or does not take over. In both domains, professional added value is shifting away from mere execution and increasingly towards understanding, interpretation, assessment, conscious steering and responsibility within AI-integrated work processes. At the same time, both sessions demonstrate that this shift does not automatically lead to stronger professional autonomy. When in-depth understanding or active steering decline, there is a risk of dependent AI use and knowledge erosion. This has direct implications for curriculum development.
The findings suggest that curriculum development should not focus exclusively on developing knowledge and skills, but also on the professional dispositions that determine whether students apply these competences critically, purposefully and responsibly within AI-integrated professional contexts.
The four scenario typologies serve not only as an analytical framework for understanding different forms of professional autonomy, but also as a (normative) frame of reference for educational development within WP3. They highlight which forms of professional autonomy programmes consider desirable in a context of structural AI integration, and which forms of AI use pose risks to sustainable competence development, critical professional practice and substantive autonomy.
The typologies do not, however, describe fixed student or professional profiles. Depending on context, task complexity, experience, time pressure and the level of AI literacy, professionals may move between different positions or combine characteristics of multiple typologies. Furthermore, both sessions make it clear that dependence is not necessarily a final state, but can also develop gradually as professionals increasingly rely on AI output whilst active reasoning, in-depth understanding and conscious steering remain less explicitly present.
At the same time, both sessions clearly position the Director as the desired professional end profile within AI-integrated professional contexts. The other typologies highlight the tensions, risks and intermediate positions that may arise when understanding, steering or professional judgement come under pressure.
| Typology | Curriculum position |
|---|---|
| Director | Intended final profile |
| Gatekeeper | Professional position with increased risk of reactive steering |
| Operator | Entry-level position with increased risk of pseudo-understanding |
| Dependent | Position to be avoided |
The Director appears in both sessions as the desired final profile of professional autonomy within an AI-integrated professional context. Students must not only be able to use AI, but also be able to consciously contextualise AI processes, actively manage them and assess their content based on an in-depth understanding of the domain and processes. Professional autonomy arises from the combination of strong AI literacy, in-depth understanding and professional dispositions that enable professionals to effectively exercise their judgement, responsibility and active steering within complex AI-integrated work processes.
The Gatekeeper represents a position in which in-depth understanding remains present, but active AI-steering is deployed to a more limited extent. Professionals possess sufficient expertise to recognise problems, anomalies or a loss of quality, but often only intervene reactively when the output proves not to align sufficiently with the professional’s intention. Both sessions highlight that this position remains vulnerable to inefficient correction loops, reactive work processes and a gradual shift towards more dependent AI use as active steering continues to decline.
The Operator represents a plausible and often natural entry point within AI-integrated education. Students often develop operational AI literacy more quickly than in-depth domain and process understanding. As a result, they are able to actively control AI systems at an early stage through prompting, workflow management and iterative adjustments, whilst their underlying understanding remains insufficiently developed to analyse output in depth, justify it professionally or resolve complex situations independently.
At the same time, the analysis reveals that this situation is particularly critical for curriculum development. Thanks to AI, students can produce professional-looking output and successfully complete assignments without the underlying understanding necessarily being equally well developed. This creates the risk of ‘false competence’, whereby active AI-steering is mistakenly interpreted as in-depth professional understanding.
The Dependent represents the position that degree programmes explicitly seek to avoid. Here, both in-depth understanding and sufficient active AI-steering are lacking, meaning that professionals largely follow AI output without sufficient substantive assessment, professional responsibility or conscious process management. Both sessions make it clear that this position is not exceptional, but can arise structurally when AI output becomes increasingly convincing, time pressure mounts and active reasoning or substantive expertise are gradually utilised less explicitly.