...
In July of 2023, OpenAI, the company behind ChatGPT, disabled public access to its AI classifier tool (Hendrik Kirchner et al. 2023). Their announcement of the change cited the classifier’s “low rate of accuracy,” even with text generated by their own ChatGPT service. While it was available, the classifier incorrectly identified human-written text as AI-generated – a “false positive” – in 9% of cases in OpenAI analysis. Independent testing found even higher rates (Elkhatat, Elsaid, and Almeer 2023).
Recent research from the University of Maryland (Saha & Feizi, 2025) finds even more concerning evidence about false positives in AI detection. Systematic testing of twelve leading AI detection tools revealed that mostdetectors failed to distinguish between different degrees of AI involvement, flagging minor edits at nearly the same rate as major rewrites. Some tools even showed the counterintuitive result of flagging lightly edited text more frequently than heavily AI-generated content.
In the context of academic integrity, the risks of false positives are significant (Klee 2023; Fowler 2023). Unreliable AI detection not only fails to improve academic integrity but may deepen existing inequalities. Non-native English speakers are flagged by AI detection tools at a disproportionate rate (Myers 2023). Other tools like Grammarly with legitimate academic applications, particularly for writers with dyslexia and other learning disabilities, also increase the likelihood of being flagged by AI detectors (Shapiro 2018; Steere 2023).
Further, recent research indicates that AI detectors exhibit bias against older or smaller LLM models, creating further inequity where students with differing levels of access to AI tools face vastly different risks of false accusations (Saha & Feizi, 2025).
| Panel | ||
|---|---|---|
| ||
For all of the reasons given above, university ITS does not currently license, support, or recommend any tool for AI writing detection. Barring a significant technological breakthrough on this front, these tools are simply not reliable enough to be incorporated into our university policies and procedures. |
...
Myers, Andrew. 2023. “AI-Detectors Biased Against Non-Native English Writers.” May 15, 2023. https://hai.stanford.edu/news/ai-detectors-biased-against-non-native-english-writers.
Saha, Shoumik, and Feizi, Soheil. 2025. "Almost AI, Almost Human: The Challenge of Detecting AI-Polished Writing." arXiv preprint arXiv:2502.15666v2. https://doi.org/10.48550/arXiv.2502.15666
Shapiro, Lisa Wood. 2018. “How Technology Helped Me Cheat Dyslexia.” Wired, June 18, 2018. https://www.wired.com/story/end-of-dyslexia/.
...