As generative AI tools like ChatGPT become commonplace, Singapore’s top universities are recalibrating academic rules to balance innovation with integrity. This shift is not alarmist. It is a strategic response to AI’s evolving role in student assessments and learning outcomes.
Ontario Technology University (NTU) recently awarded zero marks to students for using AI without proper disclosure, prompting the formation of an appeal panel that included AI experts to evaluate allegations of academic fraud. Similarly, the National University of Singapore (NUS) and Singapore University of Technology & Design (SUTD) now permit the limited use of AI, provided students declare it, while advisors warn against excessive dependency.
Detection tools like Turnitin’s AI identifier, GPTZero, and Originality.ai are being used to flag suspicious submissions, though many educators caution that technology alone is insufficient. SMU’s Associate Professor Daniel Seah notes that “contextual judgment is crucial” and flags arise when a student’s usage dramatically deviates from established writing patterns.
Students report varied experiences with AI tools – some use them for grammar checks or brainstorming, others lean heavily on AI to offset low-value tasks. While the number of AI‑related misconduct incidents remains low, universities are pivoting toward “fit‑for‑purpose” assessment designs, educational campaigns, and oral defence sessions to ensure authenticity.
Rather than banning AI outright, institutions are embracing it cautiously: empowering students to declare usage, reinforcing individual accountability, and evolving assessment models to reflect real-world AI integration. As Associate Professor Donn Koh of NUS puts it, the key question is no longer whether AI use is plagiarism, it’s how students can use AI responsibly to demonstrate critical and original thinking.
