Universities Ban AI Detectors Over False‑Positive Distrust; Some Teachers Drop Writing Assignments
| Source: Techmeme | Original article
Several universities have banned AI detection tools after faculty and students expressed distrust over false positives, while some instructors have responded by canceling writing assignments.
Universities across North America are pulling the plug on AI‑detector software after mounting complaints that the tools generate too many false positives, eroding trust between students and instructors. At several campuses, administrators have formally barred the use of plagiarism‑checkers that claim to spot text produced by large language models, citing faculty concerns that innocent work is being flagged and that the resulting disputes are consuming valuable teaching time. Professor Timothy Paustian, who has “tried everything” to keep assessments fair, is among those who have abandoned the technology altogether.
The backlash has pushed some educators to take the drastic step of cancelling writing assignments altogether, opting instead for oral presentations, in‑class essays or project‑based assessments that are harder to outsource to AI. The move underscores a broader dilemma: while institutions seek to safeguard academic integrity, the reliability of current detection methods remains questionable, and the fear of penalising students for tools they may not have used is prompting a reevaluation of assessment design.
The shift matters because it could reshape how higher education measures learning outcomes in an era where AI‑generated text is increasingly accessible. If detectors are sidelined, universities may need to develop new policies, invest in alternative verification methods, or redesign curricula to minimise reliance on written assignments.
Observers will watch whether professional bodies or accreditation agencies issue guidance on acceptable AI‑detector use, and whether vendors respond with more transparent algorithms. The next few weeks could see a wave of institutional statements, pilot programs for hybrid assessment models, and possibly legal challenges if students claim wrongful accusations based on flawed detection tools.
Sources
Back to AIPULSEN