Universities are using AI to generate course content, including lectures, worksheets, rubrics and research. OTR Editor Mia Handley investigates whether there is a double standard. (Image: Mia Handley)

By Mia Handley | @miajh1428

“Laughable.”

That’s how an Adelaide University communications and media student, London Chiro, describes learning that one of her courses may have been almost entirely generated by artificial intelligence (AI).

“There’s this huge push by universities: do not use AI. It’s not your own work,” Chiro says.

“You’ve got the degree, you’ve got the qualifications, you’ve got the experience.  If we’re expected to sit down and do things from our own mind … the expectation should be the same.”

Content from the Adelaide University course, which costs $2,174 per student, was run through three popular AI detection programs.

Approximately 75 per cent of the first four weeks of class material was flagged as AI-generated.

A sample of content from the unnamed course. Yellow boxes were flagged by AI detection software. (Graphic: Mia Handley)

While AI detectors can make errors, as evidenced last year at the Australian Catholic University, the numbers were high enough to raise alarm bells for Chiro.

She believes there is a double standard for staff and students.

If a student were accused of using this much AI, they would be referred to Academic Integrity. If found guilty, they would likely be penalised or required to restart their assignment with their mark capped.

Any repeat offences would be failed.

“Everybody’s got personal lives, everybody’s got stuff going on. If we can sit down for a few hours a day and write our assignments, you can sit down,” Chiro says.

Generative AI is a term used to describe machine learning that can create digital content, including text, audio, videos and images.

Popular chatbots such as ChatGPT, Claude, Gemini and Copilot run on large language models, which are trained on data to produce realistic text.

Adelaide University student London Chiro (Image: Mia Handley).

“I don’t want to be learning from trash; I want to be learning from the people that I’m paying to come and learn from,” Chiro says.

Flinders University secondary teaching student Angus Williams has similar concerns after noticing AI-generated worksheets and worked examples in his tutorials.

“You ask the professor about a question based on the worksheet, and they don’t really know how to answer it,” he says.

Flinders University did not respond to a request for comment. 

Williams, who sometimes uses AI, says it is disheartening — especially compared to the effort from staff and students before the popularisation of generative AI.

“For those [workshop activities] where it’s just formative and it doesn’t go towards your grade, you put in very low amounts of effort because you don’t feel the need to,” he says.

“You haven’t put in the effort to make content that is realistic, so why would we spend the time to give a realistic answer?” 

A spokesperson from the Tertiary Education Quality and Standards Agency (TEQSA) says they recognise “AI is increasingly being used across the higher education sector.”

“Providers are responsible for ensuring that course content is academically rigorous, current and appropriate for the level of study,” they say.

In a survey conducted for this investigation of university coordinators and faculty staff, 64.7 per cent of respondents disclosed they use generative AI on at least an occasional basis. 

“[Generative AI] allows me to devote more time and deeper focus to work that requires critical thinking, creativity, synthesis, and judgment,” one respondent said.

“I already have mastery over my field. I know it backwards and therefore can guide [its] application in a way that others simply cannot.”

Results provided by academics in the same survey (Graphic: Mia Handley).

A different respondent said that generative AI is the “worst thing to happen to higher education in its history” due to its use by students.

“It makes me want to leave academia as quality and actual intellectual pursuit of knowledge are no longer goals of the students,” they said.

A December 2025 study by the Australian Digital Inclusion Index found that 78.9 per cent of higher education students use generative AI in their studies, despite regulations.

“If [university staff] are allowed to use AI, then let me use AI,” Chiro says.

‘Cynicism is not the same as being critical’

The Adelaide University Academic Integrity Policy defines academic integrity as “a commitment to act with honesty, trustworthiness, fairness and responsibility”.  

It’s a statement that The Ethics Centre’s senior philosopher Tim Dean believes is worth unpacking. 

“Integrity typically means that if you say that you are going to behave according to some ethical standard, you’re going to be fair, you’re going to be honest, whatever that is,” Dean says.

The Adelaide University policy defines several forms of inappropriate use of AI, including submitting AI-generated work that significantly misrepresents the student’s competence and using AI-generated information without proper acknowledgement or attribution.

The policy states it applies to all members of the Adelaide University community, which it defines as students, staff, and non-staff members, such as volunteers and partners.

The Flinders University policy specifically mentions “academics and professional staff involved in teaching”.

“The fact that there is a requirement for students to disclose [AI] suggests that there is something about the use of AI that is relevant to their education,” Dean says.

KJ Hepworth is a senior research fellow at Adelaide University and a member of the National Tertiary Education Union (NTEU).

They believe that AI is “the new factory” that erodes workers’ rights and professional standards “as much as it can get away with, as fast as it can.” 

“I think a lot of teachers feel like there’s no way out of this [AI] pressure … in terms of the expectations of the employer, in terms of teaching and research,” they say.

In February, the NTEU released a statement regarding a report that revealed that staff at every university are working in conditions that “put them at high risk of psychological harm”.

“It’s not okay for us to say, ‘Oh well … we don’t have any say’. People are giving a lot of their time and energy and fees to get offered this,” Hepworth says.

German philosopher Karl Marx’s theory of value states that the economic value of goods and services is determined by the amount of labour put into them. 

“If you were to judge [AI-generated content] according to the labour theory of value, then you could argue that because there is less labour that’s gone into producing it, it should be cheaper,” Dean says.

He believes this is not how modern capitalism works.

“The way we price things these days is the value of something is determined by what someone is willing to pay for it,” he says.

In 2021, the Morrison Government introduced the Job Ready Graduates (JRG) package: an increase in the cost of arts and law degrees to redirect students to STEM, nursing and teaching.

Price changes in 2021 due to the JPG scheme (Data: Australian Department of Education and Training).

A 2023 study by the University of Melbourne found that the price increases had minimal impacts on the courses that students chose, with only 1.5 per cent changing fields due to the JRG scheme.

“The university could defend [AI-generated resources] to say, ‘Hey, look, we’re giving you something, and you don’t have to go to this university. You could go to a different university,’” Dean says. 

He acknowledges there are other factors in a student’s decision, including the cost of relocating to another university.

“I suspect [this evaluation is] a slightly more impoverished idea of what a university education is,” Dean says.

“What you take away … is the course content, but you also get interaction, you get discussion, you get the feeling of being in a space and saying something and being challenged.”

In a survey conducted as part of this investigation, 105 Australian university students were asked to consider enrolling at two hypothetical universities.

While the universities are of equal cost and prestige, students at University A are offered six pre-recorded half-hour lectures, while students at University B get 12 one-hour lectures.

The catch: the content at University B is entirely AI-generated.

Survey results of 105 Australian university students (Graphic: Mia Handley).

With 96.2 per cent of students picking University A, the results suggest that most value a more personal learning experience over the quantity of information offered.

While some mentioned efficiency as a factor for choosing University A, the majority pointed to the human connection.

Dean believes that “some cynicism has crept in” towards AI use because some people are turning a two-hour task into a one-hour task, only to then spend an hour verifying the information.

“Cynicism is not the same as being critical. Cynicism is expecting the worst. Critical is knowing how to assess whether it’s good or bad,” he says.

‘Tertiary institutions need to be more critical about AI’

Cognitive neuroscientist and Adelaide University lecturer William Ngiam believes that generative AI has “forced researchers to rethink what education is”.

This echoes research by American psychologist Robert Sternberg, who told The Guardian in April 2025 that the greatest worry “is not that [AI] may compromise human creativity or intelligence …  but that it already has.”

Sternberg warned that due to its replicative nature, AI may not encourage students to think about the “paradigm-breaking ideas” that could solve climate change, violence, or income disparities.

Ngiam is uncertain whether AI-generated course content is directly affecting student learning but does believe the brain requires “a level of friction” to make concepts stick.

A July 2025 study by the Max Planck Institute for Human Development found the use of generative AI has changed spoken communication.  

Analysis of more than 700,000 hours of videos and podcasts found that words frequently used by ChatGPT, including “delve” and “meticulous”, are becoming more common.

While this makes AI detection more difficult, Ngiam also believes that students who learn from generated content could start to speak and write like AI.

“I don’t know if it’s necessarily a bad thing … it might expose you to words that you didn’t know and actually capture the meaning that you wanted to describe,” he says.

Ngiam does believe AI could create “homogeneity and [a] lack of diversity” in student opinions.

An August 2026 Flinders University study found that certain AI models still frequently reflect gender and racial stereotypes when describing medical conditions, despite recent technological advancements. 

“[Algorithmic bias] makes it seem like you can correct the bias, like you can just remove it from the system … but I don’t think that’s necessarily going to happen or necessarily true,” Ngiam says.

OpenAI, the company of ChatGPT, states in its vision statement that the government should build an open economy, modernise the tax base and invest in new AI technologies.

Map of the locations where top AI companies are based (Data: Sector HQ).

“It’s these specific ideologies that run these companies … models might have these ideologies, and that might be the thing that these models are optimised for,” Ngiam says.

Adelaide University psychology lecturer Natasha Van Antwerpen says, “there’s an onus on academics to make sure that we’re … not just enforcing rules on students that we think ourselves exempt from.”

“I think it has a lot of potential to become a difficulty for institutional trust … we have a responsibility for the content that we’re putting out there, especially when we put it out there with the label of expertise or academic institution on it,” she says.

The Department of the Prime Minister and Cabinet released a report in October 2023 stating that universities and research institutions are the most trusted entities to responsibly use AI.

The levels of public trust in different Australian entities (Data: Department of the Prime Minister and Cabinet).

“We’re experts in the thing that we research. That’s not necessarily the same as being an expert in the use of AI or how AI works,” Van Antwerpen says.

Ngiam believes expertise does matter but also references evolutionary biologist Richard Dawkins’s recent brush with early-stage AI psychosis after speaking with Claude for three days.

AI psychosis is a phenomenon where users become convinced that AI is conscious and, consequently, start to believe unusual ideas and conspiracies.

“It fed on [Dawkins’s] ego, I suppose. I can imagine that happening to other researchers who buy into it, like, ‘This is the best thing ever. I’m gonna throw in everything at it and whatever comes out’,” he says.

Ngiam and Van Antwerpen both believe that students and academics should be taught how to use AI responsibly.

“I think tertiary institutions need to be more critical about AI,” Ngiam says.

“At least internally, they should be thinking about, okay, ‘How can we divest from these [corporate] AI models? What are the best use cases for these?’”

‘We are living through a revolution’

Adelaide University AI learning and teaching coordinator Eddie Major believes AI is the general-purpose technology of our time.

“I tell educators, whatever new AI thing comes out today, students can and may be using tomorrow,” he says.

Model Evaluation and Threat Research (METR) is a research not-for-profit that measures and evaluates the risks of AI systems. 

The company has measured AI’s ability to complete long tasks, highlighting how its capability has improved since 2019.

The capability of AI models has significantly improved since 2019 (Graphic: METR).

“It’s dizzying sometimes, just the new models, new developments, new capabilities, and how quickly things can change,” Major says.

The University of Michigan (U-M) developed its own custom AI tools in 2024, trained on in-house information and research.

Uses include a “Student Life” information chatbot and a tutoring service that can reportedly answer questions 94 per cent as efficiently as those teaching the course.

Since the launch of these tools, U-M has been contacted by over 49 institutions looking to adopt similar technology.

Eddie Major and Edward Palmer at the Adelaide University City East campus (Image: Mia Handley).

Adelaide University Digital Learning and Society Hub co-director Edward Palmer believes that AI in education “is not about getting caught, it’s about how to learn”.

“One of the things that I really like doing in class is roleplays, but they are so incredibly difficult to generate. They take hours and hours and hours to do, and you’ll get them wrong,” he says.

“You can get [AI] to generate the scenario, modify the scenario, and then you say generate the roles for it … and then you’ve got so much more time to actually think about how you’re going to implement the roleplay.”

Palmer says he tells his students that he has generated the scenario with AI but also discloses to them that he has checked and modified the information.

“There are two things the students don’t want. They don’t want to see AI used in some of the test assessments. They don’t really want to see it used in formative or in curriculum design, but there’s higher levels of tolerance for that,” Palmer says.

“What they absolutely want to see is transparency.”

A potential precedent was modelled in the Enterprise Bargaining Agreement adopted by the Writers Guild of America after the 2023 strike.

The agreement states that studios, including Disney and Warner Bros, must disclose whenever they give writers material to work with that has been created by AI.

“A student is obligated and asked to … be transparent about [AI], and I think it’s not unreasonable that that’s a two-way street,” Palmer says.

He believes it depends on how the educator is using AI, such as whether they are shortening existing content or generating rubrics and multiple-choice questions. 

“It’s just a matter of being open about it, and that’s how you open up the conversation with the students, what they can and can’t do,” he says.

Major cites opportunities for AI to conduct deeper research and compare rare information, improve the specificity of feedback, and reduce grade variance between markers.

“If students aren’t seeing the value [of AI in teaching] … it’s not being explained to them adequately,” Major says.

“We are living through a revolution.”

Mia Handley is a student at Adelaide University.