Australia AI Ethics Framework

Overview

The Australian Government published eight voluntary AI Ethics Principles in November 2019. They guide responsible design, development and use of AI; they are not a law that commenced in September 2024. Keep this voluntary framework distinct from applicable legislation and later policy proposals.

The eight voluntary AI Ethics Principles are: (1) Human, societal, and environmental wellbeing; (2) Human-centred values; (3) Fairness; (4) Privacy protection and security; (5) Reliability and safety; (6) Transparency and explainability; (7) Contestability; and (8) Accountability.

The government’s interim response to the Safe and Responsible AI consultation was announced on 17 January 2024. A separate high-risk mandatory-guardrails proposals paper opened on 5 September 2024. The official consultation now states that the government will not proceed at this time with those previous proposals. They must not be presented as enacted or inevitably incoming requirements.

The December 2025 National AI Plan builds on existing legal frameworks. Separately, the government announced Australian AI infrastructure and training standards in July 2026; the August National Cabinet statement described legislation intended for early 2027. These are distinct proposals, including large data-centre requirements, not a revival of the former high-risk guardrails or an already enacted general AI law.

From 10 December 2026, APP entities using personal information in qualifying automated decision-making must include specified information in their privacy policies under APP 1.7–1.9. The obligations concern the kinds of personal information and decisions involved. They do not by themselves create a general individual right to an explanation or human review. Assess the statutory significance threshold, entity coverage and any other applicable duties.

The Australian Human Rights Commission has also been active in the AI governance space, publishing a report on human rights and technology that recommended specific protections against AI-related discrimination and bias. The Commission's work has influenced the development of the broader governance framework.

Financial-services and healthcare organisations should assess the ASIC, APRA or TGA instruments applicable to their activities, along with other existing laws. Those sector-specific obligations are not created by the voluntary AI Ethics Principles. Determine the relevant entity, product and use-case scope with the accountable compliance owner.

The Office of the Australian Information Commissioner (OAIC) has provided guidance on the intersection of AI and privacy law, emphasising that existing privacy obligations apply to AI systems processing personal information and that organisations should adopt privacy by design approaches for AI development.

For an AI evaluation, separate three questions: which existing laws apply, which voluntary governance practices you choose to adopt, and which proposed or future obligations need monitoring. The Privacy Act does not automatically cover every Australian organisation or every AI interaction; assess APP-entity coverage and relevant exceptions. A security product or completed checklist does not establish legal compliance.