Australia’s AI Ethics Principles
Eight voluntary principles from 2019 for responsible AI, from human wellbeing to contestability and accountability. Not enforceable; still the vocabulary Australian guidance and tenders use.
- Status
- Voluntary · Australia (Commonwealth) · DISR
- Applies to
- Everyone and no one: voluntary, aimed at any organisation designing, developing or using AI.
- Primary source
- Official document →
- Last reviewed
- 28 September 2026
What it is
Eight principles the Australian Government published in 2019 to describe what responsible AI should look like: human, societal and environmental wellbeing; human-centred values; fairness; privacy protection and security; reliability and safety; transparency and explainability; contestability; and accountability. They are voluntary and aspirational, written for anyone who designs, builds or uses AI. They impose no obligation. They matter because every Australian document that came after, from the Voluntary AI Safety Standard to the Guidance for AI Adoption to state government AI frameworks, is built on them, and because tenders and policies still quote them by name.
What it requires
Nothing enforceable. As statements of intent: AI should benefit people, society and the environment; respect human rights, diversity and individual autonomy; be inclusive and not result in unfair discrimination; uphold privacy and secure data; operate reliably for its intended purpose; disclose responsibly so people know when AI significantly affects them; give people a timely way to challenge an AI outcome that affects them; and keep the people responsible for each phase of an AI system identifiable and accountable, with human oversight. Two of those, contestability and accountability, later became concrete elsewhere: the Privacy Act’s automated decision-making transparency rules from December 2026, and the “decide who is accountable” practice in the Guidance for AI Adoption.
Does this reach your business?
Only through other doors. No regulator enforces the principles. A tender or a customer questionnaire may ask you to show how your AI use aligns with them, and the NSW and federal government AI frameworks require their suppliers to. If you govern AI under the Guidance for AI Adoption and the Privacy Act, you meet the principles without treating them as a separate exercise.
What we recommend
Our advice is not to build a separate ethics process around these. Treat them as the vocabulary and the Guidance for AI Adoption as the checklist. When a tender asks for “alignment with Australia’s AI Ethics Principles”, answer with your record: who owns each AI system (accountability), what it may be given (privacy), who reviews its output (reliability, human oversight), whether people are told (transparency), and how someone can challenge a result (contestability). That record answers the question better than a statement of values does.
Questions people ask
No. They are voluntary principles published in 2019. Binding obligations come from existing law such as the Privacy Act and consumer law.
The Australian Government's Department of Industry, Science and Resources, following public consultation in 2019.
The guidance turns the principles into six practical steps. Follow the guidance and you are applying the principles.
Some do, especially government and large enterprise. Answer with your governance record rather than a values statement.