guide · AI and student information · last reviewed 18 October 2026
Using AI with student information in Australian schools
Each state and territory has its own privacy law and its own department rules on generative AI, and Catholic and independent schools add their own. This guide covers what applies nationally, compares the states, and links to a guide for each one.
At a glance
- Seven of the eight government school systems limit or advise against identifiable student information in AI tools; Western Australia has no public Department AI policy.
- In five, consent does not lift that limit: Victoria, NSW, the ACT, Tasmania and the Northern Territory.
- ST4S does not currently assess AI services that process health and wellbeing data; a scope decision, not an assessment outcome.
- From 10 December 2026, privacy policies must describe computer programs used in decisions that significantly affect people.
General information, not legal advice. Last reviewed 18 October 2026; rules change, so check the sources below.
The national framework
The Australian Framework for Generative AI in Schools (October 2023; its 2024 review, endorsed in June 2025, found it still fit for purpose) is a shared set of expectations, not law. Lines that matter for student information:
- Privacy disclosure: school communities are "proactively informed about how and what data will be collected, used, and shared while using generative AI tools, and consent is sought where needed".
- Protection of student inputs: staff "take appropriate care when entering information into generative AI tools which may compromise any individual's data privacy".
- Human responsibility: "teachers and school leaders retain control of decision making and remain accountable for decisions that are supported by the use of generative AI tools".
- Contestability: people affected have "opportunities to question, the use or outputs of the tools and any decisions informed by the tools".
ST4S and AI
ST4S does not currently assess AI services that process health and wellbeing data; that is a scope decision, not an assessment outcome. As at August 2026, the ST4S Excluded / High Risk list (marked "Updates in progress - 31 August 2026") also lists AI "Designed to process personal or sensitive information", AI that determines or predicts "student disability, learning difficulties", and administrative decision-making "which have a human in the loop step".
ST4S says an exclusion "does not necessarily indicate that tools or products with these features are unsuitable. It does mean that ST4S has not assessed that aspect", and recommends schools complete their own internal risk assessment. A Responsible AI standard is in pilot for suppliers that have already passed ST4S.
Federal Privacy Act changes
The federal Privacy Act covers Catholic and independent schools and their service providers. Three changes matter:
- Automated decisions, from 10 December 2026. Privacy policies must describe where a computer program makes, or does something "substantially and directly related to making", a decision that "could reasonably be expected to significantly affect the rights or interests of an individual", using their personal information.
- Children's Online Privacy Code, due by 10 December 2026. The OAIC must register it by then. The exposure draft (31 March 2026) gives "online school management systems that monitor student performance" as an example of a service in scope, and proposes parental consent for under-15s, consent lasting no more than 12 months, and a privacy impact assessment for new services. The final code may differ.
- A statutory tort for serious invasions of privacy, in force since 10 June 2025.
The OAIC's guidance on AI products (updated 17 January 2025) says privacy obligations apply to information put into an AI system and to what it produces, and that "Consent cannot be implied merely because an individual was notified of a proposed collection of personal information."
How the states compare
Government schools only. Each row links to the full guide.
| State | Department rule on AI | Parent consent | Breach notification | Approvals |
|---|---|---|---|---|
| VIC | No personal information or student audio loaded onto generative AI tools | Opt-in consent needed, but does not lift the rule | Department procedures (government); federal scheme (non-government) | PIA; ST4S; School Software Panel from October 2027 |
| NSW | No personal, sensitive or confidential information in AI tools the Department has not reviewed | Product-specific consent outside marketplaces; does not lift the rule | Mandatory since 28 November 2023 | Marketplaces or Assessed IT; Department AI approval |
| QLD | No personal information in generative AI tools not approved by the department | Valid consent for sensitive information | Mandatory for departments since 1 July 2025 | Information security assessment; FAIRA at department level |
| SA | Department advises no personal information in public generative AI tools; staff guidance needs login | No AI-specific rule found | No statute; notify the Privacy Committee as soon as possible | Department SACSF cloud and supplier risk assessment |
| WA | No public Department AI policy; AI assurance self-assessment for AI-driven systems | As the Third Party Service Risk Report sets, with opt-out | From 1 January 2027 | Catalogue or risk assessment; PIA where significant |
| TAS | No sensitive or personally identifiable information as input to any generative AI tool | No consent exception | No mandatory scheme | ITS Application Risk Assessment; endorsed list |
| NT | No personal or sensitive information in generative AI; public or unclassified only | Express consent for sensitive information and data leaving the NT; does not lift the rule | No mandatory scheme | Online and Cloud System Request; Department assessment by risk |
| ACT | No student personal information entered into generative AI | Permission forms for sign-in software; does not lift the rule | Federal scheme not generally applicable; voluntary reporting | Directorate endorsement; ACT AI Policy self-assessment |
State and territory guides
Catholic and independent schools
The ISA/NCEC Privacy Compliance Manual (April 2025) has a section on AI. It notes that AI use "may pose serious privacy risks" and suggests a privacy impact assessment "for each AI system", contracts that stop providers using personal information for other purposes such as training, specific collection notices, human oversight and labelling of AI content.
Dioceses and schools set their own rules on top. The Melbourne Archdiocese Catholic Schools staff code (2025) says "We must not enter personally identifiable information or corporate data into such tools", and Catholic Education Western Australia's AI Framework (v1.1) requires AI tools to be on its Approved Technology list. See the Catholic schools and independent schools compliance pages.
Using Descant within these rules
- AI off unless your school adds it. AI features are an optional paid add-on, and a school can switch AI off at any time. With AI off, nothing about a student is sent to an AI model, and case management, referrals, support plans, consent records and notes staff write themselves keep working.
- AI only with recorded parent consent. Where your rules allow AI, a school can require recorded consent before AI processes a student's information.
- Documents families sign online (coming). Families will be able to read and sign other school documents online, with no AI involved; consent through a secure link already works.
- Staff review every AI draft. Drafts are labelled, stay with staff, and are never sent to families automatically. Processing is in Australia.
Read how Descant keeps student data safe, our privacy policy (including how AI is used in decisions), the state-by-state compliance pages, and the note templates staff can write from with AI off.
Sources
Each source was read on 18 October 2026. Dates are those shown on the page or document. State sources are in each guide.
- Australian Framework for Generative AI in Schools (page modified 17 June 2025)
- ST4S Excluded / High Risk list (marked "Updates in progress - 31 August 2026")
- ST4S: the Responsible AI standard (read 18 October 2026)
- Privacy and Other Legislation Amendment Act 2024 (assented 10 December 2024)
- OAIC: Children's Online Privacy Code exposure draft (PDF) (released 31 March 2026)
- OAIC: privacy and the use of commercially available AI products (updated 17 January 2025)
- ISA/NCEC Privacy Compliance Manual (PDF) (April 2025)
This guide is general information, not legal advice. It describes public rules as at 18 October 2026, and makes no claim that Descant is approved or assessed by any department or by ST4S.