Currently reading Quince et al. (2025) on “Student perceptions of GenAI integration into engineering practice: How students interpret liability, safety, professional conduct and ethics across disciplines”

I’m getting a bit sick of reading GenAI papers, but I guess not reading is also not an alternative… And this one is actually very interesting!

Quince et al. (2025) use discipline-specific case studies on the use of GenAI worked on by 48 engineering students from different disciplines, to investigate how students respond to questions regarding liability, safety, professional conduct, and ethical considerations. They mapped student responses to an 32-item taxonomy of ethical considerations when using GenAI, and find several interesting things:

• All students identified “control and oversight” and “decision making, risk & uncertainty” as relevant ethical considerations, and stated that “ultimate human responsibility is non-negotiable“.
• 98% of the students (so all but one) also identified “output quality and accuracy” (“GenAI outputs require rigorous validation“) and “transparency and accountability”
• The next most common themes were “process automation” (83%), “psychological well being” (75%) and “ethical integrity” and “organisational structure” (71%)
• Five themes were not mentioned at all: “data ownership”, “enhanced user experience”, “feedback & improvement”, “power and hegemony”, and “language fluency”. “Data collection and utilisation” and “equity and accessibility” were only mentioned by one student

Interestingly, they find disciplinary differences*, with mechatronics and electrical engineering students showing the widest range of considerations, at the same time, mechanical engineering was the only discipline where students did not identify “human labour” as a consideration. It is not surprising that students that have gone through different educational paths would look at ethics of GenAI differently, especially since training (and I am assuming that that’s similar in Australia to here) isn’t coordinated, but likely depends on what individual teachers choose to highlight in their own, specific courses. So it makes sense that students focus on practical implications first, and Quince et al. (2025) explain that “students don’t develop a deep ethical understanding just by being exposed to GenAI; they build it through guided, student-active learning that makes hidden impacts visible“, and refer to other studies that suggest that those broader considerations are learnable, and that “when students are explicitly prompted to think beyond tool performance, they do articulate wider issues such as privacy, bias, and social consequences“.

What I take away from this article is (surprise!) that rather than only giving students a one-size-fits-all introduction to GenAI, we need to really embed those ethical considerations into their disciplinary context (and most likely have them taught by teachers from within that context, not by some external expert) and everyday practice. Quince et al. (2025) recommend that “educators should embed oversight by design into curricula, requiring explicit documentation of AI use, validation steps, and accountability lines within engineering assessments and capstone projects. It is also essential to broaden the curriculum with activities that expose students to data governance, equity, power dynamics, and environmental externalities to complement their existing focus on control and accuracy“.

Looks like we won’t run out of things to do in 2026 either!

*but I think here we really need to be careful and remember that all this very much depends on context and that we cannot assume that those disciplinary findings would be similar at other universities! The main point for us is that there are differences, and that there are likely disciplinary differences also at LTH, but not what exactly they are!

Quince, Z., Nikolic, S., & Goh, S. (2025) Student perceptions of GenAI integration into engineering practice: How students interpret liability, safety, professional conduct and ethics across disciplines. International Journal of Mechanical Engineering Education, 0(0), 03064190251407621. https://doi.org/10.1177/03064190251407621 

Featured image and pics below from today’s icy double dip.

When I bring the camera into the water with me, I always stay in longer than if I only dip myself… But I just love wave pics right from water level!

It was also really interesting to see how small ice pieces floated past in clusters! But it does feel weird to be in the water when all these little sharp edges touch you…

The things we do to take pictures of ice…

Some more ice, and some icicles on the stairs!