Generative AI is changing work — but how much does it change yours? Search your occupation and see what share of the tasks AI can already do, and what stays human.
Click an occupation to see the full profile. Notably: digital routine jobs are most exposed, while physical and interpersonal work is safest.
Transparent and contextual — not an oracle. It rates the typical tasks of an occupation against the current state of generative AI.
The AI-exposure score (0–100) estimates what share of an occupation's typical core tasks generative AI (text, image, code, speech, analysis) can already do or substantially accelerate today. "High" does not mean "will be replaced" — it means "changes a lot".
The assessment draws on established research on the AI exposure of occupations (incl. Eloundou et al., "GPTs are GPTs", and the IAB substitutability potentials) and applies it to the current state of generative AI and the German labour market. It rates the typical tasks of an occupation — not individual people or employers.
Exposure is not the same as replacement. For most knowledge jobs, AI accelerates the work (augmentation) — it takes over routine, while humans keep judgement, responsibility and relationships. What disappears fully is mostly narrow, purely digital routine work.
No. The score measures what share of the typical tasks AI can take over or accelerate. For most jobs that means AI makes the work faster (augmentation), not redundant. What gets fully replaced is mostly narrow, purely digital routine work.
The assessment draws on established research on the AI exposure of occupations (incl. Eloundou et al., "GPTs are GPTs", and the IAB substitutability potentials) and applies it to the current state of generative AI and the German labour market. It rates occupations, not individual people.
We cover around 100 common occupations and keep extending the list. If yours is missing, the closest occupation or the occupational field gives a good orientation.
No, an orientation. How much AI really changes a job depends on the pace of technology, regulation, cost and employers. The score shows the potential with today's generative AI — not destiny.
Status: 23 June 2026
This assessment is an orientation, not a prediction or a statement about any specific job. It rates the typical tasks of an occupation against the current state of generative AI; actual impact depends on technology, regulation, cost and employer.
We put AI where it genuinely takes work off your people, and upskill them for what remains: judgement, relationships, responsibility.