A student at the University of California Davis experienced a distressing situation when she was falsely accused of cheating on a paper using artificial intelligence. The incident, reported by Rolling Stone, unfolded when Louise Stivers, a graduating senior, had her paper flagged by the plagiarism-checking software Turnitin, leading to an investigation by the school.
Stivers expressed her anxiety and frustration, as she had to defend herself without any external support and prove her innocence. The unjust accusation added to the stress of her final semester, causing her grades to suffer. This case is another in an increasing number of examples that highlight the serious flaws in tools like Turnitin, which is increasingly misidentifying students as cheaters.
A harrowing experience, thanks to AI
The 21-year-old political science student described the incident as demotivating and a significant waste of time. She expressed that the time could have been better spent on studying, doing homework, and preparing her law school applications. Rolling Stone also revealed that Stivers was not the only UC Davis student who had been wrongfully accused of cheating due to AI-detection software.
Another UC Davis student, William Quarterman, a senior history major, faced a similar ordeal. His professor unjustly accused him of plagiarism after using the AI-detection tool GPTZero, resulting in him going through the same academic integrity audit as Stivers. Fortunately, Quarterman and his father provided guidance to Stivers, supporting her in navigating the complex academic review process at UC Davis.
Also read: ChatGPT and Bluetooth earbuds utilised for cheating in Civil Service Exam
Eventually, the university administrators acknowledged to Stivers that Turnitin’s AI detection tool was still in beta testing and they had received early access to it. The software company admits on its website that while the tool is supposedly 98 per cent accurate, there are instances of false positives. They requested that Stivers share her experience for feedback.
A blunt knife – how AI detection tools keep on failing
Tools like Turnitin and GPTZero have proven to be inadequate for their intended purpose. Recent tests showed that GPTZero fell significantly short, even OpenAI’s own detection app struggled to reliably differentiate between human-written and AI-generated text.
OpenAI also acknowledges on its website that its tool is not always accurate and should not be the sole basis for determining whether a document was generated using AI.
Unseen consequences
Although Stivers was eventually cleared by UC Davis, the investigation remains on her record, which she will need to disclose to law schools and state bar associations. The university has not apologized for the incident, leaving Stivers accountable for the software’s error, which could have negative consequences for her future career.
This situation raises concerns about the use of flawed AI plagiarism-detection tools in academia, as more students may be wrongfully accused in the future, potentially leading to severe consequences for their academic and professional lives.
Read all the Latest News, Trending News, Cricket News, Bollywood News,
India News and Entertainment News here. Follow us on Facebook, Twitter and Instagram.