Key takeaways
- From 10 December 2026, entities regulated by the Privacy Act 1988 (Cth) (Privacy Act) must include new information in their privacy policies about their use of computer programs to make or inform decisions about individuals.
- The obligation under 1.7 to 1.9 of the Australian Privacy Principles (APPs), applies wherever a computer program is used to make, or do, something substantially and directly related to making a decision, that could significantly affect an individual's rights or interests, using their personal information.
- The APPs broadly define "Computer program" to encompass its ordinary meaning and to include artificial intelligence, rule-based processes, ordinary software and generative AI tools such as chatbots.
- The OAIC has now released final guidance, including an updated Chapter 1 of the APP Guidelines, a fact sheet, a flowchart and a supplementary fact sheet for government agencies, giving organisations worked examples and practical benchmarks to apply.
- Organisations that are unsure whether a system falls within scope should take a cautious approach and disclose it.
- Aged care providers and retirement village operators using our RK DocsConnect* privacy policy templates already have a dedicated clause addressing this obligation but are required to confirm it reflects their actual practice.
Introduction
From 10 December 2026, organisations covered by the Privacy Act will need to provide further detail in their privacy policies about how they use computer programs to make decisions about people. This obligation was introduced by the Privacy and Other Legislation Amendment Act 2024 (Cth) and sits under APP 1.7 of the Australian Privacy Principles.
On 30 September 2026, the OAIC released its final guidance on the new requirement, including an updated Chapter 1 of the APP Guidelines, a fact sheet, a flowchart and a supplementary fact sheet for government agencies. The guidance reflects feedback from 90 written submissions received during the OAIC's consultation, from academia, civil society, government and industry.
What the obligation requires
An organisation is caught by the transparency obligation where all three of the following criteria is true:
- The organisation has arranged for a computer program to make, or do something substantially and directly related to making, a decision; and
- that decision could reasonably be expected to significantly affect an individual's rights or interests; and
- personal information about the individual is used in the operation of the computer program to make the decision or do the related thing.
Where all three criteria is met, the privacy policy must describe the kinds of personal information used in the operation of the computer program, the kinds of decisions made solely by the computer program, and the kinds of decisions for which the computer program does something substantially and directly related to making the decision.
The guidance also confirms that commercially sensitive information and trade secrets about a computer program are excluded from the disclosure requirement, although information that would simply be embarrassing or attract criticism is not protected on that basis.
The definition of "Computer program" under the APPs is deliberately broad. The OAIC confirms it covers pre-programmed rule-based processes, artificial intelligence and machine learning, ordinary software, apps, word-processing tools, and generative AI including chatbots. A spreadsheet with an embedded formula can be enough. So can a rules-based eligibility calculator that has been in use for a decade, with no machine learning involved at all.
The real challenge: working out what counts
Most of the practical difficulty for organisations will involve deciding what is a computer program within their automated systems. Almost every organisation now uses some form of automated process, whether that is rostering software, a CRM that stores leads, recruitment software that shortlists candidates, or a chatbot answering customer queries. The obligation is not triggered by automation for its own sake. It is triggered by the combination of automation, personal information and a decision that could significantly affect someone's rights or interests.
The OAIC has provided further guidance on the key phrases, substantially and directly. "Substantially" means the computer program is a key factor in facilitating the human decision-making. "Directly" means the program has a direct connection with making the decision. A program can meet this test whether its output is advisory or determinative, and the OAIC considers that machine learning or generative AI outputs used to make decisions that significantly affect rights or interests will generally be in scope unless subject to extensive human oversight and control. A human clicking "approve" on an AI-generated recommendation will not, on its own, take an organisation out of scope. To assist APP entities with determining whether both the substantially and directly limbs are satisfied, the APPs detail a non-exhaustive list of relevant considerations under APP 1.53.
On the significance limb, the OAIC states that the impact of the decision "must be more than trivial and must have the potential to considerably influence the circumstances or outcomes for the individual concerned". Vulnerability is also relevant. The guidance notes that a decision may have a greater impact on people experiencing vulnerability than on the general population, which is a useful reminder for sectors such as aged care, disability services and financial hardship support.
The OAIC has helpfully provided a non-exhaustive list of decisions it considers would generally be in scope, including:
- facial recognition technology used in a retail store or stadium for watchlist matching;
- recruitment software used to sort through candidate profiles and make hiring decisions;
- differential or personalised pricing practices used by online retailers in the sale of significant goods;
- computer programs used to prioritise the provision of health or disability services to individuals;
- computer programs which determine eligibility for the provision of a government benefit or housing assistance;
- computer programs which determine eligibility for admission to an education or training program;
- computer programs used to approve or reject a loan or credit application;
- programmatic advertising relating to the decision to sell significant goods or services;
- computer programs involved in assessing an individual's eligibility for insurance;
- reports created by artificial intelligence used to rank employee performance, determine promotions, or decide bonus or remuneration outcomes;
- computer programs involved in decisions relating to the provision of a scholarship, grant or other financial support; and
- computer programs used to process immigration applications or enforce border controls at an airport.
Many organisations will recognise at least one of their own systems in that list, often without having previously thought of it as automated decision making.
Where an organisation is genuinely unsure whether a particular use falls inside the obligation, the OAIC's own advice is to err on the side of disclosure. Entities "in doubt of whether their use of a computer program falls within the scope of the transparency obligation should take a cautious approach and include information in their APP privacy policy necessary to comply with APP 1.8 requirements". With that in mind, the following is what this obligation means in practice for most organisations.
What this means for organisations
These reforms are relevant to virtually every organisation covered by the Privacy Act, not just those building or procuring AI systems. In particular:
- organisations should treat the 10 December 2026 deadline as a project, not a quick policy tweak, given the broad definition of "computer program" and the fact that most organisations will have more in-scope systems than they initially assume;
- organisations should factor vendor and procurement reviews into their timeline, since the disclosure obligation sits with the organisation even where a third-party computer program is used;
- privacy policies should be drafted at a level of detail that is meaningful to a reasonable person, neither burying significant uses in generic language nor overwhelming readers with unnecessary technical detail.
Aged care and retirement living: a head start
Organisations using our RK DocsConnect privacy policy templates for aged care providers and retirement village operators already have a running start. Those templates include a dedicated Automated Decision Making and Artificial Intelligence section, with drafting options to reflect whether an organisation uses no automated decision making, or uses computer programs in a more limited, human-overseen way.
That head start does not mean the job is done. Each provider still needs to confirm which option reflects its actual practice, and populate the specific kinds of personal information and decisions involved, rather than leaving the drafting at a generic level. The relevant templates are available through RK DocsConnect.
How can we help?
Russell Kennedy’s Corporate & Commercial team can help organisations to:
- assess which computer programs are caught by the new obligation;
- review vendor arrangements and update privacy policy disclosures at the right level of detail;
- for aged care and retirement living clients, finalise the Automated Decision Making and Artificial Intelligence section of your RK DocsConnect template.
Please contact a member of our Corporate & Commercial team to discuss how this obligation affects your organisation.
If you’d like to stay up to date with Russell Kennedy Alerts and Events, you can subscribe to our mailing list here.
The information contained in this Insight is intended as general commentary and should not be regarded as legal advice. Should you require specific advice on the topics or areas discussed, please contact the Corporate & Commercial team at Russell Kennedy.
*RK DocsConnect Pty Ltd is an associated company of law practice Russell Kennedy and does not engage in legal practice or carry out legal work.