AI Safety and Regulatory Standards

Australia does not yet have a dedicated AI law. AI systems are governed instead by a mix of existing law (privacy, consumer protection, anti-discrimination, work health and safety), sector-specific regulators, and voluntary national guidance. In 2025 and 2026 the government moved away from a proposed set of mandatory guardrails for high-risk AI in favour of this guidance-plus-existing-law approach, backed by a new Australian AI Safety Institute and mandatory obligations for Commonwealth agencies.

For organisations, this means responsible AI is a self-assessed obligation, not a checklist handed down by a regulator. Boards and executives are expected to demonstrate genuine governance: who is accountable, how risks are identified and managed, and how systems are tested and monitored, using the frameworks below as reference points.

This page collects the frameworks and standards that most commonly apply to Australian organisations building or deploying AI, along with the international standards worth knowing if you operate across borders or want to benchmark against global practice. It is a reference, not legal advice. If you want help applying any of this to your organisation, see our Blueprint, Rulebook, and Safeguard services or get in touch.

Australia's AI Governance Framework

National AI Centre – Guidance for AI Adoption: Foundations

The current national guidance, published October 2025. Sets out six essential practices for responsible AI governance: decide accountability, understand impacts, measure and manage risks, share information, test and monitor, and maintain human control. Applies across the AI supply chain, developers and deployers alike.

National AI Centre – Guidance for AI Adoption: Implementation Guidance

The practical companion to the six practices, including a crosswalk showing how they relate to the original ten guardrails, legal risk examples by harm type, and alignment notes for ISO/IEC 42001 and the NIST AI RMF.

Voluntary AI Safety Standard (2024)

The original national standard, released September 2024, built around ten voluntary guardrails. Superseded by the Guidance for AI Adoption above, but still useful background on the reasoning behind the current framework.

Australia's AI Ethics Principles

Eight high-level principles (human, societal and environmental wellbeing; human-centred values; fairness; privacy protection and security; reliability and safety; transparency and explainability; contestability; accountability) that underpin all of the guidance above.

Australia's AI Safety Institute

Established in 2026 to test AI systems, assess risks, and advise government on where regulatory gaps genuinely exist.

Standards Australia – Artificial Intelligence

Australia's national standards body's work program on AI, including its Artificial Intelligence Standards Roadmap.

Privacy and Human Rights Law That Already Applies

OAIC – Automated Decision-Making Transparency (Privacy Act)

New transparency obligations for automated decision-making under Australian Privacy Principle 1 take effect 10 December 2026. Organisations using AI to make or assist decisions about individuals need privacy policies that explain this.

Australian Human Rights Commission – Technology and Human Rights

The Commission's ongoing work on AI, automated decision-making, and human rights, including guidance relevant to recruitment and other high-impact uses of AI.

International Standards and Frameworks

Useful for organisations operating across borders, selling into regulated markets, or wanting an externally recognised benchmark.