AI-adapted and assessable learning outcomes for KTH’s Master’s theses
The project involves developing KTH-common AI rules, learning outcomes, and assessment criteria for the master's degree project course, adapted to a reality with generative AI and ensuring a sufficiently high level across all degree projects.
Project context
The degree project is the single most important course at KTH and occupies a quarter of the master's degree.
An analysis conducted by Viggo Kann in 2021 (a) of KTH's current KTH common intended learning outcomes (ILO) (b) from 2015 and actual degree project learning outcomes showed that the KTH learning outcomes are vague, are not complete in terms of Master’s of Science in Engineering and are not fully examinable. Therefore, the KTH learning outcomes are not followed by the majority of KTH's degree project courses. There are many variants – today's degree project courses are therefore not equivalent.
Generative AI is affecting degree projects and, in the absence of clear rules and assessment procedures, may lead to a decline in quality. Clear rules for AI use and more process assessment (c) are needed.
(a) KTH-mål för examensarbeten, översynsbehov 2021 (Swedish ppt, docs.google)
(b) Rules for Master's thesis as of 2015-07-01 (Swedish, pdf)
(c) Hinder and detect prohibited use of generative AI (intra.kth.se)
Examples from EECS:
- Master's thesis at EECS (Swedish, docs.google)
- Master's degree project rules and routines EECS (intra.kth.se)
- Use of generative AI in the master degree project (Canvas)
Related projects within Future Education at KTH:
- Quality and diversity in assessment within programmes (2313-EECS) . In this project, new assessment criteria (e) and school-wide requirements (f) for degree projects, as well as rules for AI use (g), were developed to guide this project.
- Assessment model for degree projects: AI-generated thesis or work process? (2517-ITM)
- TP2AI: Thesis Progress Through Peers and AI (Philipp Haller EECS, end date 2026-12-31)
- Application of AI in biotechnology theses (2612-CBH)
Purpose (outcome)
The aim is to develop clearer, more reality-adapted KTH-common learning outcomes and criteria for degree projects that allow programme directors (PA) to identify shortcomings better and explain the constructive alignment between the degree objectives and the programme's course objectives.
The results from this project will affect KTH's guidelines for the Degree project for students starting studies from 1 July 2015. and can be spread to other higher education institutions through the national examination network, NEHU.
Project results (output)
The project will deliver the following concrete results:
- New KTH-wide intended learning outcomes and assessment criteria for degree projects at advanced level (for Degree of Master of Science in Engineering, Degree of Master of Science in Engineering and Teacher Education, Degree of Degree of Architect and Degree of Master of Science) such as:
- are adapted to a reality of generative AI, guarantees a sufficiently high level of all degree projects (passes a UKÄ review),
- can be examined within the framework of a degree project (basis for constructive linking),
- are so general that they fit all degree project courses and are formulated according to current guidelines and templates, as well as AI rules for the degree project;
- are well-written, well-thought-out and anchored.
- Recommended KTH-wide rules for the use of generative AI for master's degree projects.
Time plan
Start date: 2026-10-01
End date: 2027-06-30
Project documentation
If you have a KTH ID, you can read the documents when logged in:
- Poster: The project’s first poster (in Swedish) will be available at the ”Storträffen” meetup autumn 2026
- Project directive (Swedish): Projektdirektiv_2631-KTH_AI-mål-MEX.pdf