so terrible! AI plagiarism detection software keeps falsely accusing students of cheating

Webmaster’s Home (ChinaZ.com) News on June 7:In order to prevent students from using AI to write papers, some colleges and universities have adopted AI plagiarism detection software, but this may be a nightmare for some students . Because a student was falsely accused of using artificial intelligence to cheat on an essay, and then was forced to defend his name during a school investigation.

As Rolling Stone reports, an essay by Louise Stivers, a graduating senior at the University of California Davis, was caught by the plagiarism checking software Turnitin mark.

With no outside support, Stivers was forced to defend herself and prove that she did not use an artificial intelligence chatbot to write her paper, adding to the pressure on her to complete her final semester. Even her grades began to slip as a result, she argued.

This incident highlights significant flaws in tools like Turnitin and how they are increasingly being used to falsely accuse students of using artificial intelligence to cheat in class.

Stivers later learned that she wasn’t even the only UC Davis student wrongly accused of cheating based on artificial intelligence detection software. Just days before coming under scrutiny for academic integrity earlier this year, USA Today published a similar report in which history major William Quarterman said his professor was using the AI ​​detection tool GPTZero He later failed the class and was falsely accused of plagiarism. Quartermaine therefore had to go through the same process of self-evidence as Stivers.

The detection by tools like Turnitin and GPTZero seems to be inaccurate. When some users tested these two tools, they found that GPTZero was far from up to par. Even ChatGPT maker OpenAI’s own detection application failed to reliably distinguish between human-written text and AI-generated text.

Eventually, school administrators admitted to Stivers that Turnitin’s AI detection tool was in beta testing and had “early access” to it. The software company admitted on its website that despite a purported 98% accuracy rate, there were "false positives."

According to Stivers, the school has not apologized. In short, she was forced to take responsibility for the software's errors — something that could affect her career. This is a worrying reality. As academics use deeply flawed AI plagiarism detection applications, more students may be wrongly accused in the near future, which could have devastating consequences for them.

The above is the detailed content of so terrible! AI plagiarism detection software keeps falsely accusing students of cheating. For more information, please follow other related articles on the PHP Chinese website!