University college Odisee will stop requiring traditional bachelor's and graduate dissertations from the 2026-2027 academic year, replacing them with practical projects linked to students' work placements. The decision reflects a direct response to how artificial intelligence can now generate large sections of written academic work with minimal student effort.
Why dissertations are being scrapped
Ann Martin, Odisee's Director of Education, said the rise of AI tools forced the institution to reconsider what a final project actually measures. "With a traditional written dissertation, students can let AI produce large parts of the work with limited personal input," she said. The reform will roll out across all bachelor's and graduate programmes at the college, which operates campuses in Brussels, Flemish Brabant and East Flanders, by the 2027-2028 academic year.
How the new practical projects will work
Instead of a standalone thesis, students will tackle real challenges drawn from their internship or workplace. Examples include designing a school playground, developing an AI tool for a retail company, or creating construction and agricultural plans. The projects will be assessed as part of the placement activity itself, rather than through a separate final examination, making the work more integrated and harder to outsource to a machine.
Completion data and pilot results
Odisee pointed to internal data showing that around 75% of students who fail to graduate do so partly because of their dissertation. Many who fail the project do not return to finish their studies. Programmes that have already tested practice-based final projects saw higher student engagement and more directly useful outcomes for employers and organisations, according to the university college.
The change is one example of how institutions are rethinking assessment design in an era where generative AI can handle traditional written assignments - a topic explored in AI for Education training.
Why this matters for education professionals
Odisee's shift signals a practical path for programmes that want to preserve academic integrity without ignoring what AI can do. For teachers, administrators and curriculum designers, it offers a worked example of swapping a vulnerable assessment format for one grounded in workplace practice. The model also tackles a persistent bottleneck: when a final thesis becomes the single biggest barrier to graduation, replacing it with an integrated project can keep struggling students in the classroom and on track to complete their qualification.
Your membership also unlocks:
