University of Tasmania cracking down on ChatGPT use amid increase in academic integrity cases

The rise of generative AI language models has seen university and high school students using the technology to produce their work for them – but Tasmanian educators are fighting back.

Conceptually, ChatGPT is an AI chatbot or artificial intelligence that can communicate through messages with humans naturally.

The University of Tasmania is cracking down on students using AI language models like ChatGPT to help them write assignments, introducing new penalties in response to a spike in academic integrity cases.

The university has created a new breach category to cover the improper use of generative language tools and is fighting fire with fire by using AI to detect certain patterns in students’ work.

The new functionality was added to UTAS’s longstanding Turnitin system in April this year. Turnitin has previously been used to pick up primarily on plagiarism and ghostwriting in assessment tasks.

While the excessive use of AI in assignments is not considered to be a form of plagiarism, it is seen as being on a par with the engagement of a ghostwriting service or contract cheating, counting as a medium-level breach.

Offending students face potential failure or a requirement to resubmit their assignment, depending on the severity of the breach.

UTAS Deputy Vice-Chancellor Education, Professor Mitch Parsell, said the university had seen cases where entire assignments had been produced by AI.

Mitch Parsell, UTAS Deputy Vice-Chancellor Education. Picture: Supplied

“[Then there’s] cases where the student has used it appropriately but probably just hasn’t acknowledged it in the right kind of way,” he said. “And that’s more of an educative response. We still take that very seriously.”

Prof Parsell said the improper use of AI was being seen across a variety of different faculties and was most common in assessment tasks involving the use of language.

While he noted that the rise of generative language tools presented a “challenge” and could pose a “huge problem” if left unchecked, Prof Parsell said the technology also made it “a really exciting time to be an educator”, saying ChatGPT was akin to a “knowledgeable friend” or “research assistant” if used ethically.

“That’s really exciting for a state that has really low educational attainment,” he said. “This could be something that really helps the state and all of our students in ways that we haven’t thought even possible up until now.”

Academics at UTAS are exploring ways in which AI could enable them to individualise assessment tasks for each student.

UTAS has seen an increase in academic integrity cases relating to the improper use of generative AI language tools by students. Picture: Istock

Prof Parsell said the university was “moving towards different forms of assessment tasks” whereby the use of AI would not be as applicable.

It’s not just UTAS that’s clamping down on AI, either, as high schools also grapple with the issue.

State schools have barred access to ChatGPT on school devices due to the tool requiring users to be over the age of 18.

A Children, Education and Young People Department spokeswoman said some schools used Turnitin to check whether student work had been written by an AI language model, much like UTAS is doing.

“The department is working closely with other jurisdictions to adapt and evolve to challenges presented by advances in technology, as well as learning from what other states are doing in relation to these programs,” she said.

According to the spokeswoman, a “diverse mix of assessment tasks are undertaken throughout the year, which helps teachers identify authenticity”.