Computing Students’ Appraisals of and Coping with Risks from Artificial Intelligence

Sammanfattning

Artificial intelligence is increasingly integrated into education, work, public services, and everyday life, creating both opportunities and risks. The aim of this study was to examine how computing students appraise and prioritise AI-related risks, and how they reason about possible responses to such risks. The study was conducted through seven semi-structured interviews with Swedish computing students from bachelor’s and master’s programmes in fields such as software engineering, computer science, systems science, applied AI, AI and automation, and interaction design. The analysis was guided by Technology Threat Avoidance Theory and an AI risk taxonomy, while also allowing additional themes to emerge from the interview material. The study shows that participants were most concerned with immediate and everyday risks, especially overreliance, deskilling, reduced critical thinking, uncritical trust, and the effects of AI on education and professional competence. Risks related to advanced AI systems, such as loss of control or existential risk, were generally perceived as distant, unlikely, and less urgent. The findings also show that participants’ understanding of what AI is shaped which risks they considered relevant. Coping responses focused mainly on education, responsible use, verification of AI outputs, and limiting reliance on AI systems. At the same time, participants often placed responsibility for broader risk mitigation on governments, authorities, or companies, while expressing limited trust in these actors. The study concludes that computing students are aware of many AI-related risks, but that educational and structural interventions are needed to strengthen future AI risk mitigation.