Ethical Dilemma in Digital Education: Privacy, Equity and Academic Integrity in the Age of AI

Artificial intelligence (AI) is increasingly embedded in digital education, reshaping how teaching, learning, and assessment are designed and experienced. From intelligent tutoring systems and adaptive learning platforms to automated grading, plagiarism detection, and remote proctoring, AI offers unprecedented opportunities to personalize instruction, enhance efficiency, and widen access to education across diverse contexts. These technological advances, however, give rise to pressing ethical dilemma that challenge traditional educational values and human rights. Issues of privacy emerge as student data are continuously collected, analyzed, and sometimes repurposed, raising concerns around surveillance, informed consent, data governance, and the long-term security of educational records. Similarly, questions of equity highlight the risks of reinforcing or amplifying existing social inequalities, as AI tools may embed algorithmic biases, disadvantage students with disabilities, or exclude learners from underrepresented linguistic and cultural communities. Equally complex are matters of academic integrity, as generative AI blurs the boundaries of authorship and originality, while automated assessments raise concerns about validity, fairness, and the cultivation of authentic learning outcomes.This paper provides a comprehensive synthesis of these debates and develops a normative-analytic framework grounded in human rights principles and professional ethics. Through a critical examination of key use-cases—including learning analytics, AI-driven writing assistance, adaptive platforms, and surveillance-based proctoring—we demonstrate both the promises and perils of AI in education. We further propose a set of governance recommendations structured around four interrelated pillars: duty of care, proportionality and necessity, transparency and contestability, and equity-by-design, emphasizing the need for participatory oversight and continuous impact assessment. Ultimately, we argue that while AI can serve as a catalyst for more inclusive and effective learning environments, its integration must be guided by ethical safeguards that uphold privacy, foster equity, and preserve academic integrity as foundational values of education.