AI detectors wrongly flag human writing, raising concerns for universities

The issue is illustrated by the case of Lauren Jager, a chemistry student at Idaho State University, who was applying for PhD programs when she learned that some universities planned to screen personal statements with AI detectors. Although she had written her essays without using AI, several online detection tools classified them as almost entirely AI-generated. Concerned that her applications could be rejected, Jager rewrote her statement in a less polished style until the detectors assigned it a lower AI score. She was later accepted into a PhD program at the University of Utah.

Researchers say the growing popularity of generative AI has made academic integrity more difficult to assess. Unlike traditional plagiarism software, which compares submitted work against existing sources, AI-generated text often produces original wording that is harder to identify through conventional methods. As a result, universities have increasingly turned to specialized AI detection tools such as GPTZero, Copyleaks, Turnitin, Grammarly and QuillBot.

However, studies cited by Nature found that the performance of these tools varies significantly. A 2025 study reported that GPTZero correctly detected many AI-generated papers but falsely identified around 16% of human written essays as AI-generated. Another study found that several leading detectors performed better on older AI models than on newer systems, while also producing inconsistent results on genuine human writing. Even historic texts such as the U.S. Declaration of Independence were repeatedly flagged as AI-generated by some detectors.

The technology may also disadvantage non-native English speakers. A Stanford University study found that more than half of English language essays written by Chinese students before the release of ChatGPT were incorrectly labeled as AI-generated, while essays written by U.S. students were classified much more accurately. Researchers attributed this bias to differences in vocabulary and sentence complexity.

Legal and institutional challenges have already emerged. In one case, a New York judge overturned disciplinary action against a student accused of using AI after the allegation relied on an AI detection app. In the United Kingdom, the higher education ombudsman warned universities about relying on AI detectors following complaints from students, including one who argued that the software was biased against their writing style. Several universities have also declined to adopt Turnitin's AI detection tool because of concerns over transparency.

Rather than focusing solely on detecting AI, many researchers believe universities should redesign assessments to evaluate students' learning process instead of only the final written product. Some institutions and technology providers are shifting toward tools that record drafting and editing histories, allowing instructors to see how an assignment was developed.

Earlier, Qazinform News Agency reported on how students’ personality traits shape their use of Generative AI in education.