The state of AI adoption at the Australian National University
ANU governs generative AI through six institutional AI Principles approved by Academic Board in June 2023. Unlike Sydney's two-lane model, UNSW's seven-level scale, or UWA's three tiers, ANU does not publish a numbered AI assessment framework. Decisions about AI use in assessment are devolved to course convenors with college-level oversight.
ANU was one of the earliest Australian universities to reject Turnitin's AI writing detection, disabling the tool from 1 January 2024 over concerns about efficacy, bias, and transparency. The institution provides Microsoft Copilot Chat to staff and students with single sign-on, with model-training opt-outs and no full Microsoft 365 Copilot rollout publicly confirmed.
Institutional position
ANU's position on generative AI is anchored by the ANU Institutional AI Principles, approved by the Academic Board in June 2023. The six principles cover commitment to excellence and integrity as AI evolves; critical engagement with AI as a research stream; clear guidance on responsible and ethical use with discipline-specific differences; AI literacy for staff and students; appropriate access aligned with cybersecurity, privacy, safety, equity and inclusiveness; and drawing on internal expertise and cross-sector collaboration.[1]
Operational guidance for teaching is owned by the Centre for Learning and Teaching, which describes its AI for Learning and Assessment guidance as 'a living document'. The model is course-convenor discretion within college-level guidance rather than a centrally numbered tier or lane scheme.[2]
ANU's Academic Integrity Rule, last updated in 2021 (pre-ChatGPT), covers AI as plagiarism if AI-generated work is presented as one's own without attribution.[3] ANUSA's 2025 Generative AI Report found that while 75% of ANU students use GenAI, only 37% understand institutional expectations, suggesting the rule's age has produced ambiguity at the student level.
ANU disabled Turnitin's AI writing detection from 1 January 2024, citing 'significant issues regarding efficacy (including false positives), potential bias, and lack of transparency.'[4] This put ANU among the earliest Australian institutions to reject the detector outright rather than treat it as discretionary supporting evidence.
Bloom at ANU
Bloom runs on Microsoft Azure OpenAI under enterprise data handling, with prompts and responses excluded from base-model training. The data posture aligns with ANU's six institutional AI Principles on responsible use, cybersecurity, privacy, and equitable access, and respects the data-class restrictions ANU's Privacy Office applies to GenAI tooling.[1][6]
Bloom can be deployed at ANU in a day, with course-specific material ingestion and convenor-controlled student access. No new procurement, no new vendor security review, and no engineering integration is required.
AI tools at ANU
Generative AI tooling at ANU is comparatively narrow at the institutional layer: Microsoft Copilot Chat (consumer-grade, with commercial data protection), plus Library-published guidance and citation conventions. No public announcement confirms a full Microsoft 365 Copilot enterprise tenant, an institutional ChatGPT licence, an in-house AI agent, or a Studiosity subscription.
Microsoft Copilot Chat
Available to ANU staff and students via ANU uID sign-in. Inputs and outputs are not used for model training. The product is the consumer-grade Copilot Chat (with a 2,000-character per turn limit and 30 turns per conversation), not the full Microsoft 365 Copilot productivity-suite licence.[5]
ANU's Privacy Office maintains a separate generative AI and data governance position, restricting which data classes may be entered into any GenAI tool.[6]
ANU Library AI guide
The ANU Library publishes a LibGuide covering generative AI literacy, citation conventions, and teaching and learning use cases. Functions as the public reference rather than a tool-licensing endorsement.[7]
AI research at ANU
ANU's institutional AI research footprint is dominated by the School of Cybernetics, founded in 2021 by Distinguished Professor Genevieve Bell. The School built on the 3A Institute (3Ai), launched in September 2017 as a partnership with CSIRO Data61. The School of Cybernetics frames its work around the safe scaling of AI in human, technical and ecological systems.[8]
Genevieve Bell served as ANU Vice-Chancellor from 1 January 2024 and resigned on 11 September 2025. After 12 months of study leave, she will return as a Distinguished Professor in the School of Cybernetics.[8]
The Humanising Machine Intelligence (HMI) Grand Challenge, launched in 2018 and led by Associate Professor Seth Lazar (philosopher), spans philosophy, computer science, social sciences and law, and was funded at over $1.5M per year for at least three years from launch. Public events on the HMI site go quiet after 2022; current operational status is not publicly clear.
Sources
- 1.ANU Learning and Teaching, ANU Institutional AI Principles (June 2023)
- 2.ANU Learning and Teaching, AI for learning and assessment
- 3.Australian National University, Guide for students: best practice when using Generative AI
- 4.TELT @ SMP, ANU, AI writing detection now available in Turnitin
- 5.ANU Learning and Teaching, How to access and use Microsoft Copilot
- 6.Australian National University, Generative AI and Data Governance
- 7.ANU Library, Artificial Intelligence including generative AI
- 8.Australian National University, School of Cybernetics
If we've got something wrong. This page reflects publicly available information as of 1 May 2026. If you work at ANU and there is something we should correct, please get in touch.