This article was co-written by Hope Van Rossen.
On 18 May 2026, the Office of the Australian Information Commissioner (OAIC) released an Issues Paper seeking stakeholder feedback to inform its forthcoming guidance on the automated decision-making transparency obligation in Australian Privacy Principle 1 (APP 1) . The obligation, introduced by the Privacy and Other Legislation Amendment Act 2024 (Cth) (POLA Act), will commence on 10 December 2026, and the OAIC intends to release its final guidance by September 2026.
Submissions on the Issues Paper close on Monday 15 June 2026.
In short, the new obligation (the ADM obligation) requires entities to include information in their privacy policies about the kinds of personal information they use, and the types of decisions they make, using “computer programs” that could “reasonably be expected to significantly affect the rights or interests of an individual”. To formulate its initial positions, the OAIC has drawn heavily on the Explanatory Memorandum to the POLA Act.
The reform brings Australia broadly in line with comparable international jurisdictions, although it is less onerous than (for example) Article 22 of the EU/UK General Data Protection Regulation (GDPR) or Principle 10 of Western Australia's Privacy and Responsible Information Sharing Act 2024, which confer additional rights such as contestability, notification and human intervention.
Who is affected?
This is a significant development in privacy law for entities across the Australian economy that use technology in their operations.
We have considered the types of business processes which are often automated, or which are candidates for automation, and which usually use personal information. These are the types of processes that deserve attention. Our thinking is that the breadth of this obligation is likely to be wide, and to illustrate the point we have formulated a high-level example list of the types of processes that APP entities must examine, broken down by sector below.
| Sector | Processes |
| All |
|
| Government | Eligibility for programs, licence granting and revocation (all tied to information about individuals relevant to the criteria underlying a decision. |
| Financial institutions | Credit applications, loan applications, “know your customer”, hardship processes, and insurance underwriting and claims processing, which are highly automatable and involve personal information including financial information. |
| Consumer markets | Loss prevention, customer complaints and loyalty processes may be affected, to the extent that they use personal information (eg complaint history, purchase history). |
| Technology | Ad-tech, cryptocurrency, gaming and outsourcing providers should scrutinise key processes involving personal information, particularly where a process uses personal information and may have substantial adverse effects (eg loss of money). |
| Life sciences and healthcare | Treatment decisions and clinical trial participation are potential areas of focus if personal information is used to determine inclusion or exclusion. |
Application of the statute
The ADM obligation applies to APP entities: this includes Commonwealth agencies and private sector organisations with a turnover above $3m regulated by the Privacy Act 1988 (Cth). From 10 December 2026, an APP entity must include the prescribed disclosures in its privacy policy where all three of the following criteria are met:1
- the entity has arranged for a computer program to make, or do a thing that is substantially and directly related to making, a decision;
- the decision could reasonably be expected to significantly affect the rights or interests of an individual; and
- personal information about the individual is used in the operation of the computer program to make the decision, or to do the thing that is substantially and directly related to making the decision.
For these purposes, “making a decision” and “doing a thing” each extend to refusing or failing to make a decision or do a thing, and a decision may affect rights or interests whether the individual is adversely or beneficially affected. Where the obligation is triggered, the entity's privacy policy must disclose the kinds of personal information used in the operation of the relevant computer programs, the kinds of decisions made solely by the operation of such programs, and the kinds of decisions for which a thing substantially and directly related to making the decision is done by the operation of such programs.
What are the key issues the OAIC seeks to resolve?
The Issues Paper canvasses the OAIC’s preliminary thinking on a number of key terms that determine the practical reach of the ADM obligation. The OAIC invites stakeholder feedback on each.
Arranged for
The words “arranged for” in APP 1.7(a) make clear that the obligation falls on the entity that arranges for the computer program to make or assist a decision, even if a different entity actually operates it. The OAIC’s examples make it clear that AI systems procured from a vendor, and use of AI chatbots by staff, are covered. In contrast, an entity that merely develops, hosts or maintains infrastructure for ADM used by another entity may not be covered. Entities are accordingly expected, during and after procurement, to actively identify, assess and oversee how third-party products and services use ADM.
Computer program
The OAIC considers that “computer program” should be construed broadly, consistent with the Explanatory Memorandum, which contemplates that the term “is intended to take its ordinary meaning and encompass a broad range of matters, including pre-programmed rule-based processes, artificial intelligence and machine learning processes to make a computer execute a task”.
In the OAIC's view, this captures a wide range of technologies including commonly used software, apps, word-processing tools, and generative AI tools (whether for text, image, video, code or other synthesis, including chatbots).
In NRF’s view, this would likely also capture some forms of smart contracts and electronic gaming markets.
Substantially and directly related to making a decision
APP 1.7(a) narrows the field of covered computer programs to those that do a thing “substantially and directly related to” making the decision. Drawing on the Explanatory Memorandum, the OAIC proposes that “substantially” means the program is a key factor in facilitating the human decision-maker’s decision, and “directly” means there is a direct connection with making the decision. The OAIC uses the example of a pre-programmed Excel formula used to score and triage domestic-violence hotline callers to illustrate its interpretation:
- if the score is a key factor in a human deciding the order in which to attend calls, this would be both substantially and directly related to making a decision; or
- if the score is instead used merely to calculate an age in years from a date of birth entered into the spreadsheet, this may be “directly” but not “substantially” related to making a decision.
The OAIC seeks views on the relevant factors for this assessment, including the degree of reliance on the ADM output, the ability and likelihood of human override of an ADM decision, the advisory or determinative nature of the output, transparency and explainability, and the integration of ADM into the decision-making workflow.
Significantly affect rights or interests
APP 1.9(c) confirms that effects on the rights or interests of an individual can be adverse or beneficial. The OAIC draws on comparable concepts in administrative law (procedural fairness in respect of impacts on livelihood, reputation, property and employment) and European materials, and notes that effects must be more than trivial but may be heightened for vulnerable individuals such as children.
APP 1.9(d) lists examples such as:
- decisions under legislation to grant or refuse a benefit to an individual; • decisions affecting an individual’s rights under a contract; and
- decisions affecting access by an individual to a significant service or support. The OAIC is consulting on what factors increase the likelihood of a decision affecting an individual’s rights or interests, what classes of person are “vulnerable”, and what amounts to a "significant service or support".
The Explanatory Memorandum gives examples including statutory benefit decisions, decisions about life insurance contracts, decisions affecting access to healthcare, and the use of computer programs to target individuals with content or advertising where this results in differential pricing for the provision of, or access to, significant goods or services, or limits employment opportunities.
Making a decision
Although the term is not defined, the OAIC looks to analogues in the Administrative Review Tribunal Act 2024 (Cth), the Administrative Decisions (Judicial Review) Act 1977 (Cth) and other statutes, each of which casts “decision” broadly to include making, suspending, revoking or refusing to make orders, certificates, licences or determinations, imposing conditions, and doing or refusing to do any other act or thing.
One key area of consultation for the OAIC appears to be whether decisions should be captured if they are such that the affected individual is not aware of, and may never become aware of the decision. A fictional case study asks whether an algorithmic ad-targeting system that effectively excludes a female graduate from receiving a job advertisement should be characterised as a “decision” for these purposes. This case study should be carefully considered by entities in the advertising and technology space.
Key questions for organisations
With commencement on 10 December 2026 and final OAIC guidance expected by September 2026, APP entities should be using the period between now and the end of 2026 to prepare in earnest. Legal counsel, risk teams, and boards should be asking questions such as:
- Can we identify all of the “decisions” we make that impact “rights” and “interests”?
- Can we find every business process that uses a “computer program” that triggers the application of the ADM obligation?
- Can we distinguish between decisions involving ADM that are caught by the amendments, and those that are not?
- Do we know how outsourced service providers have implemented AI for our organisation which might then trigger a requirement to comply with the ADM obligation?
What's next?
Practical steps include:
- Map ADM use across the business. Entities should identify all computer programs that make decisions, or do things substantially and directly related to making decisions, where personal information is used and where the decision could reasonably be expected to significantly affect an individual's rights or interests.
- Review third-party and procurement arrangements. Because the obligation attaches to entities that have “arranged for” ADM, entities should review upstream arrangements to identify ADM that has been embedded in third-party products and services, and put in place appropriate monitoring and oversight throughout the procurement lifecycle.
- Update privacy policies. Privacy policies will need to be revised to disclose where personal information is used in the operation of a computer program, the kinds of decisions made solely by ADM, and the kinds of decisions for which ADM does a thing substantially and directly related to making the decision.
- Consider interaction with other frameworks. Entities (particularly Commonwealth agencies) should consider how the ADM obligation interacts with their other obligations (including State and Federal AI frameworks, anti-discrimination law, administrative law review rights, and sector-specific regulatory frameworks).
What must my privacy policy include if I’m covered?
The OAIC’s preliminary view is that disclosures should be:
- clearly articulated in plain language;
- structured to enable consumers to seek further information;
- sufficiently specific to be meaningful without overwhelming detail;
- logically organised; and
- framed so that the information and the decision can be challenged or contested.
How Norton Rose Fulbright can help
Our privacy team can help you prepare for the ADM obligation, including by:
- developing tailored questionnaires and undertaking due diligence to help map ADM use across business units and third-party vendors;
- advising on whether particular ADM use cases are caught by APP 1.7;
- reviewing and refreshing privacy policies to incorporate ADM disclosures that meet the OAIC's expectations; and
- advising on procurement, contractual and governance frameworks for third-party ADM.
Footnotes
-
Privacy and Other Legislation Amendment Act 2024 (Cth) clause 88 (new APP1.7).