Ethical Implications of Generative AI in Education: Risks, Practices, and Governance (Research Proposal)

Abstract

Large language models and multimodal generators as generative artificial intelligence (GenAI) are quickly changing the teaching, learning, and assessment. Although GenAI will potentially offer personalized feedback, scalable tutoring, and content generation, it carries with it very pressing ethical implications: the risk to academic integrity, biased and mis-informed content, data protection and privacy concerns, deskilling of teachers, and unequal access. This mixed-method research will (1) map ethical risks as perceived by students and faculty, (2) through measurement assess prevalence and drivers of problematic GenAI use, (3) through assessment institutional policies and pedagogical practices informing ethical use. The results will be used to support a series of evidence-based recommendations regarding the ethical implementation of GenAI in curriculum and assessment. The research is based on higher learning facilities in three countries to be able to gather because of cross-cultural difference.

Background & Problem Statement

In the year 2022, the application of generative AI, such as ChatGPT and other LLM, has become common in higher education. With their usage, they can render students more productive and efficient in accessing learning materials, but recent evidence suggests that they bring in serious ethical dilemmas: students can apply GenAI unethically (as in fail to acknowledge their work), automated evaluations may reproduce bias, and institutions have failed to effectively respond (by placing ad-hoc ban, vague policy, or limiting teaching) (Giannakos, 2024; Hasanein et al., 2023). The policies will be futile or unjust without structural empirical studies between the stakeholders and situations.

Literature Review & Theoretical Framework

Research Questions & Objectives

Primary question

What are the principal ethical risks posed by GenAI in higher education, and which institutional and pedagogical responses are most effective and equitable?

Objectives

1. Characterize ethical harms (integrity breaches, bias, privacy risks, teacher deskilling).
2. Quantify prevalence, drivers, and student/faculty attitudes toward GenAI use.
3. Evaluate the effectiveness of current institutional policies and pedagogical mitigations.
4. Produce practical, context-sensitive governance guidelines.

Methodology

Design: Explanatory sequential mixed methods (quant → qual → policy analysis).

Quantitative phase (survey):

Qualitative phase

Semi-structured interviews with 30 faculty, 20 administrators (academic integrity officers), and focus groups with students in order to investigate ethics perception and impacts of policy. Thematic coding will be used to define ethical issues which arise and contextual documents.

Policy & document analysis:

Strategic analysis of institutional policies (n=60) and assessment design practices to categorize governance strategy (ban, permissive, integrative, educative). The comparison will be evaluated based on the accordance with the best practices presented in the literature (Kovári et al., 2024).

Ethics & Data Protection

The informed consent, data anonymization and institutional review board (IRB) approvals. The safe-data protocols will be applied towards sensitive inputs of GenAI tools into the research (no personal data of students uploaded to third-party LLMs).

Expected Contribution & Significance

This research will (a) offer solid empirical data on the risk of ethics and motivation of GenAI abuse, (b) trace what the institution can do to respond (and what has intensified the inequalities), and (c) generate evidence-based and practical regulations that instructors and leaders can follow to implement GenAI ethically. Findings could be used by policymakers and accreditation agencies to come up with reasonable and practical standards.

Timeline (12 months)

Months 1–2: Literature review, IRB approvals, survey instrument piloting.
Months 3–5: Survey distribution and quantitative analysis.
Months 6–8: Interviews, focus groups, qualitative analysis.
Months 9–10: Policy/document analysis and synthesis.
Months 11–12: Writeup, stakeholder workshops, dissemination.