GenAI and Academic Integrity: Ethical Challenges, Practical Approaches, and Responsible Strategies

A widespread challenge in the era of using GenAI is determining whether users are substituting required effort with AI tools, leading to a misleading sense of accomplishment. This issue extends to the difficulty of identifying whether students’ work is predominantly AI-generated, and teachers’ concern of becoming “plagiarism detectors”. Concerns also arise regarding the degradation of critical thinking skills, and effects on task-specific self-confidence. What can be done to respond to these issues is to move from a ‘GenAI’ paradigm to an ‘Agentic AI’ paradigm: Clear expectations on AI use, including course level and departmental/institutional level policies and example-setting is paramount. Second, focusing on detection and punishment may undermine trust between learners and teachers. Instead, education on responsibility, capability, limits and ethics may be more effective. Lastly, educating students about data privacy is vital, as public AI tools are not secure; using an internal LLM/AI is the best practice.