Typology 3: The Operator
(High steering + limited understanding)
The Operator is situated in the quadrant comprising high steering and limited understanding. It is a position in which professionals manage AI actively, consciously and often with technical proficiency, whilst their underlying understanding of the domain and processes remains less developed. The typology highlights that strong AI literacy and active AI-steering do not automatically go hand in hand with in-depth substantive understanding.
What characterises the Operator?
The professional manages AI actively and purposefully. Targeted prompts are formulated, workflows are set up, output is adapted and processes are iteratively adjusted. The Operator often possesses a considerable level of operational AI literacy and knows how AI systems can be put to practical use to quickly generate usable output.
At the same time, their underlying understanding of the domain and processes remains relatively limited. The professional can use AI output operationally and actively steer it, but does not always possess a sufficiently in-depth understanding to reconstruct the rationale behind decisions, justify them professionally, or analyse complex situations independently.
The distinction from the Gatekeeper lies primarily in the degree of active steering. Whereas the Gatekeeper allows AI to function relatively freely and intervenes mainly reactively, the Operator constantly attempts to actively steer the AI. The distinction from the Director, on the other hand, lies in the underlying understanding. The Operator actively steers, but sometimes lacks the in-depth conceptual basis on which that steering is substantively grounded.
Why is this a high-risk position?
The Executor position primarily entails risks due to the tension between active steering and limited substantive understanding.
Professionals often feel they have a strong degree of control over the process because they actively steer the AI, refine prompts and constantly adjust the output. In predictable and routine situations, this approach generally works well enough. However, as soon as contexts become more complex or unexpected problems arise, it becomes apparent that the underlying understanding is sometimes not sufficiently deep to analyse deviations in terms of content or to justify them professionally.
Furthermore, active AI-steering easily creates a sense of control that does not always correspond to the actual depth of understanding. The professional constantly makes adjustments and actively takes decisions during the process, but this does not necessarily mean that the underlying functioning, quality or implications of the output are fully understood.
As a result, in day-to-day practice, the Operator is often difficult to distinguish from the Director. Both actively control AI, produce output and make decisions during the work process. The difference usually only becomes apparent at moments when in-depth understanding is required: when systems fail, when choices need to be justified in substance, or when complex professional judgements have to be made.
What do the sessions reveal?
Within the CMM session, the Operator manifests itself in the active use of AI for brainstorming, content creation, variant development and iteration. Participants describe how AI rapidly generates ideas, images and content, which are then actively refined, adapted and combined.
The professional therefore clearly takes an active role in steering the process, but the underlying assessment is often based on experience, intuition and instinct. The session reveals that participants pay a great deal of attention to output, speed and practical applicability, whilst the explicit articulation of underlying assessment criteria remains more limited.
Within the IT session, the same logic emerges among developers who make intensive use of AI during implementation, for example for boilerplate code, functions or proposed solutions. Developers actively adjust the output, integrate generated code into existing workflows and refine solutions iteratively.
At the same time, the session explicitly highlights the risk that developers may accept code simply because it ‘works’, without fully understanding the underlying logic or broader architectural implications. As a result, the vulnerability often remains hidden as long as systems function stably, but becomes apparent during debugging, maintenance, complex integrations, performance issues and security vulnerabilities.
Implications for the curriculum
The ‘Implementer’ role highlights that operational AI literacy and active AI-steering are, in themselves, insufficient for professional autonomy. Professionals may be technically proficient at managing AI systems, organising workflows efficiently and actively refining output, whilst their underlying understanding of the domain and processes remains insufficient to correctly assess the quality, reliability or implications of that output.
This makes this position particularly vulnerable in educational contexts. Thanks to AI, students can successfully complete assignments, produce professional-looking output and carry out complex tasks without the underlying understanding necessarily being equally well developed. The active steering of AI can give the impression that competences have been mastered, whilst in reality students remain heavily reliant on AI output to arrive at solutions.
Within IT, this can lead to code that functions operationally but is not sufficiently understood to be maintained, debugged or justified architecturally in a safe manner. Within CMM, this can result in campaigns or content that appear technically convincing but, in terms of substance, do not sufficiently align with the target audience, context or brand positioning.
The Operator therefore represents one of the most critical areas of tension for curriculum development within AI-integrated education. The central challenge is shifting from assessing end products to making underlying understanding, professional reasoning and substantive decision-making visible.
At the same time, the Operator is a plausible entry point within AI-integrated education. Students often develop operational AI skills more quickly than in-depth professional understanding. This enables them to start using AI actively and purposefully at an early stage, whilst the subject-matter expertise on which professional judgement is based is still very much a work in progress.
The central educational challenge therefore lies not only in teaching students to use and manage AI systems, but also in gradually guiding them towards a deep understanding of the subject area and processes. The evolution from Operator to Director requires not only greater technical proficiency, but above all an increase in subject-matter insight, professional judgement and responsibility.