Beyond the Benefits: A Systematic Review of the Harms and Consequences of Generative AI in Computing Education

TL;DR

This study systematically reviews harms and unintended consequences of Generative AI (GenAI) in computing education, covering research published from 2022 to 2025. Using a PRISMA-guided literature review across ACM Digital Library, IEEE Xplore, and Scopus, the authors screened 1,677 references to 224 included papers and performed deductive coding on harms, tool types, tasks, populations, and evidence levels. They identify six broad harm categories—cognition, metacognition, assessment/academic integrity, equity, social, and instructional/logistical—with cognitive and metacognitive harms most prevalent, though many harms lack strong empirical backing and social harms are under-reported. The paper offers a taxonomy of GenAI harms tailored to computing education and discusses grand challenges and a roadmap for future work to maximize benefits while mitigating harms, informing educators, researchers, and tool developers.

Abstract

Generative artificial intelligence (GenAI) has already had a big impact on computing education with prior research identifying many benefits. However, recent studies have also identified potential risks and harms. To continue maximizing AI benefits while addressing the harms and unintended consequences, we conducted a systematic literature review of research focusing on the risks, harms, and unintended consequences of GenAI in computing education. Our search of ACM DL, IEEE Xplore, and Scopus (2022-2025) resulted in 1,677 papers, which were then filtered to 224 based on our inclusion and exclusion criteria. Guided by best practices for systematic reviews, four reviewers independently extracted publication year, learner population, research method, contribution type, GenAI technology, and educational task information from each paper. We then coded each paper for concrete harm categories such as academic integrity, cognitive effects, and trust issues. Our analysis shows patterns in how and where harms appear, highlights methodological gaps and opportunities for more rigorous evidence, and identifies under-explored harms and student populations. By synthesizing these insights, we intend to equip educators, computing students, researchers, and developers with a clear picture of the harms associated with GenAI in computing education.

Figures (7)

• Figure 1: PRISMA diagram detailing our identification, screening, and assessment of paper eligibility for inclusion.
• Figure 2: Stacked bar chart showing the percentage of papers that discussed harms of GenAI in computing education. Greyed bars represent papers not analyzed. Papers from 2025 and other excluded works are omitted.
• Figure 3: The stacked bar chart shows the proportion of papers based on their highest level of evidence. Papers from 2025 are omitted from this visualization
• Figure 4: Umbrella harms (e.g., Metacognitive) are bolded and represent the total number of instances across all associated sub-harms. Because sub-harms are not mutually exclusive, a single paper may be counted multiple times within an umbrella category if it includes multiple sub-harms with the same type of evidence.
• Figure 5: Harms as a percentage of each population. 69.6% of instances involved introductory students, 12.0% upper-level undergraduates, 9.3% graduate students, and 9.1% educators