How the University of Cologne integrated AI into teaching with clear educator control
From a transparent pilot at the Institute of Genetics to broader adoption across multiple teaching areas.
About the institution and contact person
As generative AI became part of daily university life, two questions appeared at the same time: how should institutions use it, and who stays in control. For Dr. Anja Neuber at the Institute of Genetics, the second question had to be solved first. She was willing to introduce AI in teaching, but only under clear conditions. Students should be able to trace where answers come from. Responses should be based on course materials. Instructors should remain able to configure the system and pause AI functions when needed.
“I can recommend acemate with full confidence because it is transparent, controllable by me and I retain control over my materials.”
Dr. Anja Neuber
Institute of Genetics, University of Cologne
Control was built in from day one
The pilot started in winter term 2024 in Genetics. The key objective was not simple activation, but a clear framework. In acemate, instructors define what topics the tutor may cover, which tone it should use and when AI functions should be paused. Answers include source references to specific documents and pages. This keeps interactions transparent for students while preserving educator responsibility for course content.
“I can recommend acemate with full confidence because it is transparent, controllable by me and I retain control over my materials.”
Initial hesitation turned into routine usage
As in many higher education settings, students were cautious at first. Any additional tool introduces new routines. The rollout was therefore positioned as an offer within course practice, not as a mandate. Short recurring learning series supported adoption. After a few weeks, initial hesitation developed into steady usage. Students who had used acemate through one semester asked to keep it in the next.
From user to product co-creator
During the pilot, Dr. Neuber became a recurring contributor to product development. Over multiple semesters, concrete classroom feedback informed feature decisions. This is especially visible in aggregated AI chat insights and in course-level topic overviews. Both functions help instructors identify content priorities and comprehension gaps early without reviewing individual student conversations.
From one pilot to multiple departments
The pilot evolved into regular use. Beyond Genetics, additional teaching areas now use acemate, including Geography. The deciding factor was not a single feature, but trust in a transparent and configurable framework. Colleagues were able to review the configuration options themselves and evaluate fit for their own teaching context.