Ethical Use of Artificial Intelligence in Engineering Education: A Systematic Review

Synthesis: Osunbunmi and colleagues (2026) present a PRISMA-guided systematic review of empirical studies on the ethical use of AI in undergraduate engineering education, moving beyond conceptual discussions of AI ethics to examine how ethical guidance is actually enacted in instructional practice. From 2,158 records they retained 99 empirical studies, coding them for core ethical principles and identifying seven recurring forms of ethical guidance: transparency and disclosure of AI use; faculty–student accountability and human oversight; designing for student independence and agency; privacy and data protection; Academic Integrity and authorship; fairness, equity, and bias mitigation; and beneficence. The review finds a consistent pattern — ethical AI guidance in engineering education is predominantly student-facing and compliance-oriented (centered on academic integrity and disclosure), while reciprocal accountability for faculty AI use and institutional responsibility remains comparatively underdeveloped — and calls for a more balanced, practice-oriented approach to responsible AI integration.

From policy to practice

The review responds to a shift in conversations about AI in engineering education from how AI can improve educational processes to how ethical governance shapes its responsible use. Prior reviews often synthesized policy documents, conceptual frameworks, or literacy interventions; this study instead focuses specifically on empirical studies of AI use in undergraduate engineering classrooms, using deductive coding of core ethical principles combined with inductive thematic analysis.

Method

Following PRISMA 2020, the authors searched eight databases (APA PsycInfo, ProQuest, ERIC, Engineering Village, Education Research Complete, ScienceDirect, Scopus, Web of Science). Inclusion criteria required empirical, peer-reviewed studies of AI use in undergraduate engineering education (2000–2025) that addressed ethical considerations explicitly or implicitly. The database search yielded 2,158 records; after deduplication and blind screening, 99 studies were retained for ethical analysis.

Seven recurring forms of ethical guidance

Thematic analysis identified seven recurring forms of ethical guidance in engineering education:

1. Transparency and disclosure of AI use — openly communicating when and how AI is used.
2. Faculty–student accountability and human oversight — taking responsibility for decisions and outcomes.
3. Designing for student independence and agency — preserving students' own Problem Solving and critical thinking.
4. Privacy and data protection — safeguarding sensitive educational data.
5. Academic integrity and authorship — ensuring authenticity of student work.
6. Fairness, equity, and bias mitigation — treating learners equitably and avoiding AI bias.
7. Beneficence — acting to do good and avoid harm.

Across the 99 studies, accountability (n = 47) and transparency (n = 34) were the most emphasized core aspects, while autonomy (n = 12) and beneficence (n = 10) were the least emphasized; integrity (n = 49) was prominent enough to be treated as a distinct theme.

A student-facing, compliance-oriented imbalance

The central finding is that ethical AI guidance in engineering education is predominantly student-facing and compliance-oriented — often centered on academic integrity and disclosure requirements — while reciprocal accountability for faculty AI use and institutional responsibility remains underdeveloped. The authors interpret this as a limitation: because engineering decisions directly affect public safety, infrastructure reliability, environmental Sustainability, and societal Well-Being, ethical AI use in engineering classrooms is a matter of professional formation and societal responsibility, not merely academic policy. They document the need to make ethical guidance more balanced across students, faculty, and institutions, and to support Educational Development so educators can guide responsible AI integration.

What this means for practice

• Faculty developers. Move training past integrity rules: across the 99 studies, accountability (n = 47) and transparency (n = 34) dominated while reciprocal accountability for faculty AI use and institutional responsibility stayed underdeveloped, so educators need to model disclosure and oversight, not only enforce it.
• Instructors. Design for student independence and agency, the least-emphasized dimension (autonomy n = 12) — require students to verify AI-generated outputs and explain their reasoning instead of submitting unreviewed AI work, which protects problem-solving and critical thinking.
• Instructors. Make transparency reciprocal in your own course: document in the syllabus and teaching materials when and how you use AI, and pair student disclosure requirements with verified, explained use.
• Administrators. Treat ethical AI use in engineering as professional formation tied to public safety, infrastructure reliability, and societal well-being, and fund Educational Development so ethical guidance is balanced across students, faculty, and institutions rather than resting on student-facing compliance.

Limitations

• The central finding describes the literature, not classroom behavior: the review codes what 99 published studies say about ethical guidance and cannot verify how that guidance is enacted in practice.
• Retention was narrow: 2,158 records yielded 184 studies using AI in engineering education and 99 that addressed ethics explicitly or implicitly; conceptual papers, opinion pieces, reviews, and non-English publications were excluded.
• No quality appraisal or study weighting was applied to the included studies, so a methodologically weak study counts as heavily as a rigorous one.
• Inclusion was restricted to English-language, peer-reviewed, empirical studies published between 2000 and 2025, so the corpus skews toward English-language higher education contexts.

Citation

Osunbunmi, I. S., Moyaki, D., Fakiyesi, V. O., Dunmoye, I. D., Nwanua, M. I., Oloyede, M., Bamidele, B. R., Feyijimi, T. R., & Hunsu, N. (2026). Ethical Use of Artificial Intelligence in Engineering Education: A Systematic Review. ASEE Annual Conference & Exposition, Paper ID #52865.