The AI disruption in engineering education: an analysis of changing student norms through cultural historical activity

Synthesis: Leino Lindell & Stöhr (2026) use Cultural-Historical Activity Theory (CHAT) to analyze how generative AI disrupts and transforms established norms and practices in engineering education from a student perspective. Thematic analysis of interviews with 25 students at a technical university in Northern Europe identifies four themes of challenge/transformation: (1) the self-directiveness of students, (2) the objectives of learning, (3) the role of the teacher, and (4) ethical aspects. Students are developing new implicit rules for using GenAI to enhance skills and understanding, driven by contradictions between traditional academic tools and new expectations for self-directed, AI-mediated learning.

Key Findings

• Four themes of GenAI-driven norm change. The analysis surfaced challenges or transformations in: students' self-directiveness, the objectives of learning, the role of the teacher, and ethical aspects of academic work.
• New implicit rules for GenAI use. Students are developing informal, emergent norms (implicit rules) for when and how to use GenAI to support their skills and understanding — norms shaped by the CHAT interplay of individual and cultural-historical context.
• Contradictions as the driver. Norm changes are driven by contradictions between traditional academic tools/practices and new expectations for self-directed, AI-mediated learning. CHAT conceptualizes these as systematic tensions that prompt individuals to question and deviate from established norms.
• A relational view of norms. Norms are treated as historically and culturally embedded implicit rules of the activity system, not individual preferences — reframing student GenAI use as a systemic, norm-shifting phenomenon.

What this means for practice

• Instructors. Assign tasks that resist easy generalization and demand meaningful learning, since students' new implicit rules treat generalizable work as legitimate territory for GenAI; the authors single this out as the mitigation for misuse risk.
• Instructors. Make the new norms explicit rather than assuming traditional ones still hold: negotiate with students what self-direction, learning objectives, the teacher's role, and ethical use now mean in this course.
• Instructors. Protect the interaction that GenAI displaces — students reported turning to it instead of their teachers for generalizable tasks — by building in checkpoints where questions have to reach a person.
• Learners. Expect to be assessed on understanding rather than output, and check with the course what counts as acceptable GenAI use; the students in this study were writing their own informal rules precisely because formal ones were unclear.
• Faculty developers. Support staff to meet the new expectations students bring, because the contradiction driving norm change includes teachers' limited capacity to respond to more frequent, self-directed support demands.

Limitations

• The study draws on qualitative interviews with 25 engineering students at a single Northern European technical university, and only the student perspective is represented — educators, program heads and educational administrators were not interviewed, and the authors state the findings may not fully translate to other contexts.
• It captures perceptions, not learning: because GenAI was used informally and outside institutional design, the authors state they cannot determine whether students' learning was actually enhanced.
• The theoretical frame is bounded: CHAT offers no conceptual tools for explaining why students select one tool over another, and the analysis did not examine cases where implicit and formal rules align.
• The temporal scope is a snapshot, so behaviors and tensions may shift quickly as AI tools develop.

Citation

Leino Lindell, T., & Stöhr, C. (2026). The AI disruption in engineering education: An analysis of changing student norms through cultural historical activity theory. Journal of Computing in Higher Education.