SINGAPORE, Aug 16 — Nanyang Technological University (NTU) will stop using its institutional artificial intelligence detector from 2027, saying such tools are “fundamentally unreliable” and cannot be used to prove academic misconduct.
The university will deactivate the detector at the end of this semester on December 31 and will no longer treat AI probability scores as evidence of cheating, Singapore-based news outlet CNA reported, citing an email to faculty members signed by associate vice-provost for education transformation Tan Seng Chee.
Tan said the tools lacked “empirical validity” and could not distinguish between unauthorised AI use and responsible use of the technology as part of the curriculum.
Instead, NTU will move away from automated surveillance towards assessment methods that document how students develop and complete their work, including standardised disclosures of AI use.
Faculty members have also been told that automated AI scores or flags should not be presented as evidence of violations, while ambiguous detection reports can create unnecessary administrative burdens and strain relationships between teachers and students.
The move comes as NTU expands access to institutional AI tools for faculty and students, with the university saying continued use of automated detection alongside sanctioned AI use would send mixed signals.
NTU’s decision follows controversy over alleged AI misuse by three students in 2025, who disputed accusations and raised concerns over due process.
NTU’s own guidance on AI detectors warns that they can produce false positives and false negatives, be bypassed through minor changes to text and potentially disadvantage students with non-native writing patterns.
The university’s shift reflects a broader debate among higher education institutions over whether AI detectors can reliably distinguish human-written work from AI-generated text.
