The CHASS President, Frank Bongiorno, delivered this presentation at the 7th annual gathering of the Canadian Australian Partnership for Open Scholarship (CAPOS), Re-Defining Open Social Scholarship in an Age of Generative ‘Intelligence’, on 2 December 2025 at the Australian National University

AI, at least in its ChatGPT garb, probably first came to the notice of most Australian academics a couple of years ago as a shiny new student cheating machine. Their initial instinct was to take steps to try to prevent their students from using it to cheat. They often discovered that ChatGPT invented references. Phew! That was a relief. I mightn’t pick it up every time, they reflected, but at least I have something to go on.

AI still invents references. We recently had a senior Australian judge complaining that one of his jobs these days is to pick up lawyers who have relied on AI and therefore cited fictional precedents. Woops. A major consultancy firm hired by the government also put fake references in their very expensive report – fakery discovered by an academic when he read the document, not by the client. Woops again.

I recall the early days. You’d look a student in the eye and tell them that their paper showed every sign of having been written with an AI tool. They’d deny it vehemently. Where did you get these references, you’d, ask, pointing to an entry in the bibliography on Nicholas Nickleby attributed to Karl Marx (I jest, but you get the idea.)

We came up with solutions. Bring back the exam, we said. That will fix it. Unfortunately, with Covid still in the air, my own university – the ANU – was still not allowing us to set invigilated examinations. But in any case, it didn’t feel right: many of us had been educated with more creative forms of assessment, and we felt we had flourished via the opportunities for research, writing and creative thinking that they had offered.

It is a mark of the ingenuity of Australian academia that we have largely moved beyond these kinds of defensive reactions. Largely, I say, because we continue to worry over AI’s capacity to allow students to cheat their way through a degree. The web is crawling with advertising reels telling you of the latest tool that will write your thesis for you. I usually turn them off in the first three seconds – so offensive do I find them to everything I think universities should stand for.

AI is also causing the organisation of which I’m president, the Council for the Humanities, Arts and Social Sciences, other concerns. We seek to represent not only the HASS fields in universities, but in the wider society, and that means we are interested in matters such as the imposition of Australian content quotas for streaming services, which we support, and the theft of creatives’ intellectual property, which we oppose. We were therefore dismayed by a Productivity Commission report which suggested that letting AI companies mine content owned by others for free would be a great boon to productivity. We said in our submission:

"The Council for the Humanities Arts and Social Sciences (CHASS) shares the wider concerns of the cultural sector about the concept of a ‘fair dealing’ exception to cover Artificial Intelligence (AI) model training. The unauthorised use of copyrighted material in any circumstances is a form of theft. In the context of a cultural sector where wages are notoriously low, and with the Australian Government’s stated policy to improve the remuneration and career pathways of creatives considered as workers, this idea seems especially misplaced. …
Commissioner Stephen King, one of the Interim Report’s authors, was recently quoted as acknowledging that ‘not everyone will be a winner … There will be people who will lose their jobs because of this technology and those people need to be looked after’."

We do not believe that … writers and artists whose interest in their own work is legally protected by Australian copyright law should be framed as casualties of progress who merely need ‘to be looked after’. These are people entitled to have their property and their creative personhood protected.

All I have said thus far no doubt marks me as a Luddite who needs to get with the program. In fact, my sense is that the perhaps excessive defensiveness of some early academic reaction to AI has largely dissipated, and we have seen some of the same ingenuity that kicked in when Covid hit now being applied to AI. For the purposes of this discussion, I am conflating education and research as ‘scholarship’ for the very good reason that they are best thought of this way.

So, where are we now?

First, university teachers are devising forms of assessment which recognise that students will be living as citizens and workers in a world where AI not only exists but becomes more powerful and ubiquitous. They need to learn how to use it critically and ethically. We have creative university teachers – historian and French Studies academic Alexis Bergantz of RMIT University is an example – guiding students towards the creation of their own research questions via deep reading, asking students to feed those questions into an AI tool to create an AI-generated essay, then inviting reflection on what the AI tool made of the question in light of the reading and reflection the student has already undertaken to get them to this stage. You can see what is happening here: the focus is on assessing process, rather than outcome. The student is learning of the capacity and limitations of the AI tools. They are practising their critical thinking. They are reflecting and writing.

Second, we know that our colleagues are using AI creatively and intelligently to deal with large amounts of data and to help translate it into new and original scholarship. Only a fool would reject this kind of process. It would be like telling mathematicians to stick to a slide rule, economists to cling to the abacus. But it is worth reflecting on what is going on here. Most scholars will come to AI with a deep sense of context. They know what questions to ask. They are able to treat the products of their work with AI critically because they have a deep understanding derived from their experience – which will usually be founded on traditional forms of humanities scholarship such as that very old-fashioned exercise of reading books and articles.

Finally, I want to make a point about a very positive result of the AI revolution that is sweeping over the humanities, arts and social sciences. It has invited us to think more carefully about what is distinctive about these fields.

One established way of thinking about the humanities, in particular, has been to focus on the acquisition of cultural capital. It is doubtful that this was ever what mattered most, but in an era when I can find out who really wrote Nicholas Nickelby, as well as a find a good plot summary and much else, on a phone in my pocket, we might need to refocus our sense of value.

But on what? Here are some suggestions.

The humanities are about emotional engagement and grappling with otherness. They take us out of ourselves and provide opportunities for connection with people who are different from us, who lived in past times, or who live in distant places, or who are just unlike us in terms of their identity, opportunity and experience. I needn’t elaborate on how important these capacities are in this era of rising racism and rampant populism. AI might be able to assist here, but it would need to be highly controlled, curated kind of assistance, drawing on the kinds of contextual knowledge I have already referred to. In the end, though, an AI tool does not know what it means to be human. We all do.

Then, there is creativity and imagination. AI can imitate creative practice. But it does not itself engage in creative practice or imaginative play. Generative AI is ‘a pattern recognition tool’. It is world of predictions and probabilities. But creative work in the humanities is about surprise, delight and even quirkiness, the connections that the creative mind can make between this, that and the other thing. The humanities are, at their best, like great poetry or art: an imaginative stretching of an audience’s sense of the meaning of words, images, forms and sounds so that they can see the world anew. Writing – as well as other forms of practice in HASS – matter here. Humanities scholars do not ‘write up their research’. Their writing is integral to the development of their understanding and to the expression of the creativity.

There is also originality. AI scrapes the digital realm; humanities scholarship takes in the whole world, including the vast domain that remains buried in the dusty archive and in the hearts and minds of real human beings – who have not told their stories in Reddit or Facebook but will do so face to face with a skilled and ethical interlocutor.

The originality and distinctiveness of humanities scholarship resides, in large part, in its status as cultured and humane practice rooted in personhood and place. Here, AI might be a useful tool. It must never become the master builder.