Australia - National AI Action Plan

Overview

The National AI Action Plan (June 18, 2021) sets a whole-of-government, economy-wide policy direction to promote the development and adoption of trusted, secure and responsible artificial intelligence. The Plan commits AUD 124.1 million in targeted measures including establishment of a National Artificial Intelligence Centre (NAIC), four AI and Digital Capability Centres, co-funded AI pilot projects, the Next Generation AI Graduates Program and a regional AI grants program. The document aligns with Australia’s broader Digital Economy Strategy and builds on Australia’s AI Ethics Framework and consultations that took place through the Techtonic summit process. Rather than creating prescriptive regulation, the Action Plan emphasises capacity building, practical guidance for industry (especially SMEs), dedicated research and scholarship funding, and coordination across government and research institutions to promote beneficial AI outcomes for Australians.

Definitions

For the purposes of the Action Plan, AI is framed broadly to include systems that perform tasks conventionally associated with human intelligence through models, algorithms and data-driven approaches. The Plan reuses definitions consistent with Australia’s AI Ethics Framework and related government guidance: "AI system" (software or service incorporating machine learning and inferencing), "adoption" (implementation by businesses and public services), "capability centres" (hubs providing access to experts, tools and training), and "responsible AI" (development and use that is fair, safe, transparent and respects rights). The Plan intentionally does not attempt to create a single technical standard definition for regulatory purposes but instead focuses on functional descriptions to support policy and program delivery.

Governance and Institutional Framework

Governance under the Action Plan is collaborative and distributed. The Department of Industry, Science and Resources (DISER) is the central policy coordinator and manages program delivery and consultation. The National Artificial Intelligence Centre was established within CSIRO’s Data61 as the operational hub to convene industry, research and government partners and to deliver guidance and tools; it subsequently became a core national capability under government stewardship. Sectoral regulators (for example, the Office of the Australian Information Commissioner for privacy, the Therapeutic Goods Administration for medical devices, and the Australian Competition and Consumer Commission where competition/consumer issues arise) retain statutory roles over area-specific risks. The Action Plan also foresees partnerships with universities, research bodies and industry to run capability centres, pilot programs and scholarship initiatives. Governance emphasizes voluntary standards, guidance and capacity programs rather than immediate legal mandates, with mechanisms for cross-agency coordination and stakeholder engagement to iterate on policy.

Key Focus Areas

The Plan sets out four principal focus areas: (1) Developing and adopting AI to transform Australian businesses — supporting SMEs via Capability Centres, pilot programs and funding to reduce barriers to entry; (2) Creating and attracting world-class AI talent — through scholarship programs, graduate placements and an expanded talent pipeline such as the Next Generation AI Graduates Program; (3) Using AI for national challenges — government–industry pilots and cross-sector projects to apply AI in health, disaster resilience, agriculture and manufacturing; and (4) Making Australia a leader in responsible and inclusive AI — reinforcing Australia’s AI Ethics Principles, supporting standards development, improving public trust and ensuring inclusion and fairness. Practical deliverables include NAIC resources (guidance, directories, navigator tools), AI adoption support for SMEs, co-funded regional grants (Catalysing the AI Opportunity in our Regions), and investments in scholarships and workforce development. Emphasis is placed on outcomes that create jobs, lift productivity and ensure AI systems reflect Australian values and human rights considerations.

Implementation Framework

Implementation relies on a mix of centrally funded programs and distributed delivery: NAIC coordinates guidance and industry engagement while Capability Centres and university partners deliver hands-on support to SMEs. Program governance includes competitive grant rounds, partnership agreements for pilots, scholarship award processes and the development of practical tools (for example, the AI Impact Navigator and voluntary safety guidance). The Action Plan sets timelines for initial roll-out of major measures (establish NAIC, stand up capability centres, launch scholarships and regional grants) and indicates that program participation will be subject to terms and reporting. Implementation also includes ongoing consultations—policy design is adaptive, using evidence from pilots and engagement to inform potential regulatory or standards work where gaps are identified.

Monitoring and Evaluation

The Action Plan commits to monitoring program delivery and measuring outcomes such as AI adoption rates among SMEs, number of graduates/scholars supported, successful pilot outcomes addressing national challenges, and the delivery of guidance and standards. NAIC and DISER are expected to maintain metrics, publish progress updates and iterate on measures based on evidence. Evaluation leverages program reporting, regular stakeholder consultation and alignment with broader national indicators under the Digital Economy Strategy. The Plan signals further consultation for responsible AI measures and potential future regulatory options should voluntary and capability measures prove insufficient to manage identified harms.

Penalties, Liability, and Appeals

The Action Plan itself is a strategy and does not create a new penalty regime. Instead, it clarifies that existing statutory frameworks remain applicable — e.g., the Privacy Act 1988 enforced by the Office of the Australian Information Commissioner, consumer protection and competition laws enforced by the ACCC, and sectoral regulators (Therapeutic Goods Administration, ASIC, etc.) that may take enforcement action where AI systems give rise to regulated harms. Program funding and grants administered under the Plan include contractual conditions and performance requirements; failure to meet those can result in funding withdrawal and contractual remedies. Where liability arises from AI-enabled decisions, existing civil liability frameworks and appeals mechanisms apply, and the Plan signals coordination across agencies to explore legal and policy gaps.

Relationship to Other Instruments

The Action Plan builds on and intersects with several earlier and concurrent instruments: Australia’s AI Ethics Framework and the Australian Human Rights Commission’s technology and human rights reports, the Government’s Digital Economy Strategy (2030), sectoral regulations (privacy, health device regulation, consumer and competition law), and CSIRO/Data61 research roadmaps. It references adoption of voluntary standards and the development of guidance tools, and it complements sector-specific initiatives such as health technology assessments and infrastructure resilience programs. The Plan is intentionally complementary to existing law and standards activity rather than replacing statutory instruments.

International Alignment

The Action Plan explicitly recognises the importance of international engagement and standards alignment. It identifies collaboration with international partners, participation in global standard-setting bodies, and alignment with global best practice as priorities to ensure Australian AI capabilities and regulatory approaches are interoperable with other jurisdictions. The Plan encourages engagement with multi-lateral fora and standards organisations to influence international norms and to import proven frameworks that protect rights while enabling innovation.

Implementation Timeline

Sources and References