Australian Voluntary AI Safety Standard: requirements, deadlines and compliance actions
Voluntary AI Safety Standard (Australia)
What is the Australian Voluntary AI Safety Standard?
The Voluntary AI Safety Standard gives Australian organisations ten guardrails for developing and deploying AI safely and responsibly: accountability and governance, risk management, data governance and protection, testing and monitoring, human control and intervention, user transparency, contestability, supply-chain transparency, record keeping, and stakeholder engagement. The guardrails were designed to align with the mandatory guardrails proposed for high-risk settings so that early adopters would be prepared if those became law.
Who does it apply to?
Voluntary; aimed at organisations deploying AI in Australia, especially deployers of AI systems in higher-risk settings. It creates no legal obligations.
Any Australian organisation deploying or developing AI (voluntary), with practical focus on deployers.
When do the requirements apply?
Published 5 September 2024 alongside the mandatory-guardrails proposals paper.
What must organisations do?
Nothing mandatory. Adopters implement the ten guardrails proportionately to the risk of their AI use cases.
Voluntary Establish accountability processes and a risk-management process (guardrails 1 and 2) Guardrails 1 and 2
Guardrail 1 asks organisations to set up accountability processes including governance, internal capability and a strategy for regulatory compliance; guardrail 2 asks for a risk-management process to identify and mitigate risks across the AI lifecycle.
Practical action: Adopt an AI policy, assign an accountable owner and run per-use-case risk assessments.
Evidence examples: AI accountability and risk-management documentation
Framework mapping (original, editorial): ISO/IEC 42001:2023 Clauses 5 and 6; NIST AI RMF 1.0 GOVERN and MAP
Obligation page Source-linked
Voluntary Test and monitor systems, enable human control, and be transparent with users (guardrails 4 to 6) Guardrails 4, 5 and 6
Test AI models and systems before deployment and monitor them in operation; enable meaningful human control and intervention; and inform end users about AI-enabled decisions, interactions with AI and AI-generated content.
Practical action: Publish AI disclosures, define override paths and keep test and monitoring records.
Evidence examples: Test reports and user disclosure records
Framework mapping (original, editorial): NIST AI RMF 1.0 MEASURE and MANAGE functions
Obligation page Source-linked
Voluntary Provide contestability, supply-chain transparency and records (guardrails 7 to 9) Guardrails 7, 8 and 9
Establish processes for people impacted by AI to challenge use or outcomes; be transparent with other organisations across the AI supply chain about data, models and systems; and keep and maintain records to allow third parties to assess compliance.
Practical action: Add AI clauses to supplier contracts and keep an AI system register with evidence links.
Evidence examples: AI system register and supplier disclosures
Framework mapping (original, editorial): NIST AI RMF 1.0 GOVERN 6.x
Obligation page Source-linked
Penalties
None; voluntary.
Official sources
-
Voluntary AI Safety Standard
Department of Industry, Science and Resources · 5 Sep 2024 · Tier 1 source
Change history
Record version 1: Initial structured record.. Full edit history is in the GitHub repository.
Frequently asked questions
- Is the Australian Voluntary AI Safety Standard mandatory?
- No. It is voluntary guidance from the Department of Industry, Science and Resources, aligned with ISO/IEC 42001 and the NIST AI RMF.
Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.