China set to outlaw use of chatbots to write dissertations

Legislation under consideration in China would allow for degrees to be revoked if students are found to have used artificial intelligence tools such as ChatGPT in the drafting of their dissertations.

The draft law on academic degrees specifies that qualifications may be revoked if any of the following is found in a dissertation or practical project: “plagiarism, forgery, data falsification, artificial intelligence ghostwriting, and other academic misconduct”. If such cases arise, universities’ degree evaluation committees would be charged with ruling on the matter.

Current regulations covering academic degrees in China date back to 1980. If the law is passed, it would most likely be the first piece of national legislation to directly address the use of chatbots in education.

Although ChatGPT is not officially available in China, many people in the country bypass restrictions or use other generative AI based on large language models.

Liu Deliang, a professor of law at Beijing Normal University, said the proposed change “highlights the challenges that artificial intelligence has brought to higher education”. He noted that existing regulations already covered other types of ghostwriting, so the main difference is that the new law “mentions this scenario explicitly”.

Since the launch of ChatGPT last year, academics have scrambled to redraw assessments to reduce the risk of students asking the chatbot to write an essay for them, while plagiarism detection firms have raced to create tools to catch cheaters.

But fail-safe solutions have proved to be in short supply, and some academics have preferred to keep an open attitude, believing that the application of AI to learning can stimulate innovation.

“A practical issue here is how to define the difference between ‘artificial intelligence assistance’ and ‘artificial intelligence ghostwriting’, which is a challenge for the degree evaluation committees and relevant detective technologies,” Professor Liu said.

“Another issue is how to balance blocking and dredging, as the old saying goes. Blocking is preventing academic misconduct with compulsory methods, but dredging requires the education authorities and universities to think about guiding students to use AI correctly.”

Possible discussions include whether research tasks such as literature reviews should be distinguished from writing, how universities can adopt and improve AI detection technologies, and whether there should be clear guidance telling students when AI may be used, Professor Liu added.

karen.liu@timeshighereducation.com