Deakin University reveals benefits, risks of generative artificial intelligence to students and staff

Generative artificial intelligence is making it harder for universities to make sure students succeed in their course, but may improve research projects. TAKE OUR POLL

Deakin University has outlined the potential benefits and risks of generative AI to students and staff in a submission to a federal inquiry.

Generative artificial intelligence is making it harder for universities to make sure students meet their course requirements, but is streamlining research work, Deakin says.

The university made a submission to a federal inquiry into the use of generative AI in the Australian education system, noting Deakin was “forging ahead” with its work in the space.

In its submission, the university said the way AI would “embed” into the everyday was going to “rapidly evolve”, with benefits and risks not yet fully known.

It said it was essential students were able to deal with “ambiguity and uncertainty” because of the obscure origins of AI outputs.

The university said while the use of generative AI for learning was not banned, it was making it “considerably harder” for educators to make sure students were meeting learning requirements.

To combat this, Deakin said its advice to staff included having open conversations about the technology, reviewing assessment criteria to take account of AI, specify to students when was appropriate use of AI, design assessments to celebrate the unique achievements of students, and develop students’ digital literacies.

“Deakin encourages staff and students to develop their awareness, knowledge and skills to use these technologies ethically and responsibly as digitally fluent citizens,” the submission said.

Despite the uncertainty, the university said there was also an “obvious potential” for researchers to harness the technology to streamline their work, while emphasising the importance of effective risk management.

It said “bespoke” specialist generative AI tools like Consensus, Elcit.org, and Scite.ai could be promoted to the academic community, as long as clear guidelines were set.

“We believe Deakin’s guidance on the use of commercial generative AI products for various aspects of research is one of the first, if not the first, of its kind in Australia,” it said.

“We regard developing the high-level principles and expectations contained in the guidance as essential to supporting our researchers to realise the benefits and avoid pitfalls (of using AI in their work).”

The university said it was also a leading contributor to the development and applications of AI through its Applied Artificial Intelligence Institute.

Eleven researchers from Deakin’s Centre for Research in Assessment and Digitial Education (CRADLE), affiliated with the Research for Educational Impact (REDI) Strategic Research Centre, also made a submission.

The researchers gave four recommendations, including:

– Fund social sciences research into how generative AI is altering knowledge making;

– Counter ethical risks through regulation of governance and policies;

– Teach and assess critical digital literacies to students and educators across all curricula; and

– Promote educational experiences that promote open horizons and allow students to flourish.

Originally published as Deakin University reveals benefits, risks of generative artificial intelligence to students and staff