Australian Universities Consider Secure Assessments While Facing Criticism Over Consultation and Resources

The Australian National University released a consultation paper to academic and teaching staff that presents three options for addressing generative AI in assessments, The Guardian reported. One option would classify assessments as secure, meaning completely free from AI cheating risk, or insecure. Another would require students to declare at which stages they used AI.

An ANU academic described the university’s approach as a “hysterical” and “reactionary response,” saying the proposals could reverse progress on inclusive assessments for students with disabilities or caring responsibilities. The same academic noted that their faculty had already issued draft guidelines ahead of the second semester that push for greater on-campus assessment, with teaching due to begin in less than two weeks and little training provided.

A second ANU academic said the sector-wide process had been “panicked,” aggravated by insufficient resources to support teaching staff.

The 2025 Australian Digital Inclusion Index found that 78.9 percent of secondary and tertiary students were using generative AI. ANU law professor Will Bateman said combating AI infiltration across Australian universities was essential to prevent the country from “shipping our national intellectual capability” to companies in California and China.

He added that securing assessments did not require returning all students to exam halls but instead demanded serious investment in fully secure devices.

The University of Queensland began implementing new policies to secure assessments earlier this year and classifies them as secure or open. Deputy vice-chancellor Kris Ryan said the university started considering responses to AI in 2023 and that secure assessments could still permit some AI use provided students demonstrate critical appraisal of the tool’s output.