Australia - National AI Strategy (2025)

Overview

The Australia National AI Plan 2025, published by the Department of Industry, Science and Resources on 2 December 2025, sets a national strategy to harness AI for economic growth, social benefit and public-sector improvement while managing risks through an adaptive, sectoral approach. The Plan is organised around three goals — capturing the opportunity, spreading the benefits and keeping Australians safe — and identifies nine concrete actions spanning infrastructure, capability, investment, adoption, skills and safety. The Plan is intended to be implemented across Commonwealth portfolios, in partnership with states and territories, industry, unions, researchers and civil society. The public release (see National AI Plan page and the official PDF at National AI Plan (PDF)) emphasises a pragmatic use of existing laws and regulators, while creating new capabilities such as an Australian Artificial Intelligence Safety Institute (AISI) and strengthening the role of the National Artificial Intelligence Centre (NAIC).

Definitions

The Plan adopts established international terminology: "AI system" follows the OECD definition used in the publication ("a machine-based system that ... infers ... how to generate outputs such as predictions, content, recommendations, or decisions"). It distinguishes between development, deployment and adoption phases of AI, and notes differing levels of autonomy and adaptiveness after deployment. The Plan refers to "high-impact" or "high-risk" uses in contexts such as healthcare, medical device software, automated decision-making affecting government services, biometric systems and other settings where harms to rights, safety or livelihoods may arise. It also references Indigenous data sovereignty frameworks where use of First Nations data requires specific governance and community engagement. These definitions guide the Plan's risk-based, sectoral approach to oversight and intervention.

Governance and Institutional Framework

The Plan assigns stewardship and coordination roles across several existing and new institutions. The Department of Industry, Science and Resources acts as lead steward for whole-of-economy strategy and will work closely with the National Artificial Intelligence Centre (NAIC) to convene industry, support SMEs and provide guidance for adoption. The Department of Finance / Digital Transformation Agency supports GovAI and whole-of-government digital standards. The Australian Public Service Commission (APSC) and Department of Finance coordinate the APS AI Plan to ensure safe government use, including mandatory capability uplift and appointment of agency Chief AI Officers (see APS AI Plan statement of intent). The Plan establishes a new Australian Artificial Intelligence Safety Institute (AISI) to provide independent technical analysis, monitoring, safety testing and advice to regulators; the AISI is intended to become operational in early 2026 and to join international networks of safety institutes (AISI announcement). Regulatory responsibilities remain with existing portfolio regulators (OAIC for privacy, ACCC for consumer law, ACMA/eSafety for online safety, TGA/Department of Health for medical devices), coordinated through whole-of-government mechanisms and the Digital Platform Regulators Forum where appropriate.

Key Focus Areas

The Plan sets out nine actions and associated focus areas: (1) build smart infrastructure — expand high-speed connectivity, data centre principles and support sovereign capability; (2) back Australian capability — invest in research, support industry scaling and strengthen the NAIC; (3) attract investment — review and align federal/state incentives and remove barriers to capital; (4) scale AI adoption — provide targeted supports to SMEs, not-for-profits and regional communities; (5) support and train Australians — expand micro-credentials, TAFE courses and lifelong learning pathways; (6) improve public services — embed secure GovAI tools, pilot generative AI in schools and uplift APS capability; (7) mitigate harms — advance AI safety science and apply regulatory tools to address online harms, deepfakes and AI-facilitated abuse; (8) promote responsible practices — voluntary standards, guidance and impact assessment tools for industry; (9) partner on global norms — international engagement and regional leadership. The Plan highlights targeted reviews in healthcare, medical device software and application of consumer and competition law to AI contexts and stresses Indigenous data governance and equity in adoption.

Implementation Framework

Implementation is designed to be iterative and evidence-driven. The Plan identifies responsible agencies for actions and establishes coordination mechanisms across Commonwealth portfolios and with states and territories. The NAIC will lead industry engagement, provide guidance (including the recently published Guidance for AI Adoption) and host tools such as the AI Adoption Tracker. The AISI will perform technical safety testing and share findings with regulators to inform proportionate interventions. The Plan signals a mix of measures: direct investment programs (grants and NAIC-led support), capability-building (training and APS mandates), policy and guidance (standards, templates, voluntary frameworks) and targeted legislative amendments where sectoral regulators identify gaps. Monitoring will draw on ABS, Jobs and Skills Australia and NAIC metrics to track adoption, skills and harms. The Plan sets out responsibilities but does not, itself, create new criminal offences or civil penalties: enforcement remains via existing statutes enforced by portfolio regulators, with the government reserving the right to propose targeted legal changes where necessary.

Monitoring and Evaluation

Progress under the Plan will be monitored through a flexible evaluation approach using national and sectoral indicators. The government will rely on the Australian Bureau of Statistics, Jobs and Skills Australia, NAIC tools (AI Adoption Tracker, Responsible AI index) and regulator reporting to measure outcomes including investment flows, SME adoption rates, workforce skilling indicators and incidence of AI-related harms. Regular reporting and periodic public updates are envisaged to maintain transparency and enable iterative refinement. The APS will report on GovAI uptake and agency compliance with the APS AI Plan, while AISI outputs will feed into regulatory assessments. The Plan anticipates stakeholder engagement, including unions, industry and civil society, to collect feedback and surface implementation challenges.

Penalties, Liability, and Appeals

As a strategic document, the National AI Plan itself does not set new, Plan-specific criminal or civil penalties. Instead, it confirms the government will rely primarily on existing sectoral laws and statutory regulators to enforce obligations and seek legislative updates where gaps are identified. Privacy breaches remain subject to the Privacy Act 1988 and OAIC enforcement; consumer harms will be addressed under the Australian Consumer Law via the ACCC; online harms and non-consensual deepfakes are addressed under the Online Safety Act and criminal law reform programs; health and medical-device risks are handled by the Therapeutic Goods regulatory framework. The Plan emphasises clear accountability lines for public-sector AI use, mechanisms for review and appeals under existing administrative law, and the role of regulators in investigating and taking enforcement action where harm or non-compliance occurs.

Relationship to Other Instruments

The Plan is explicitly designed to complement and build on existing Australian instruments: the APS AI Plan (AI Plan for the Australian Public Service), the Policy for the Responsible Use of AI in Government, Data and Digital Government Strategy, Online Safety Act 2021, Privacy Act 1988, Australian Consumer Law, Therapeutic Goods Act/regulation for medical devices, and Indigenous data governance frameworks (e.g. Framework for Governance of Indigenous Data). It also refers to prior government initiatives (National Reconstruction Fund, Research and Development Tax Incentive) and existing NAIC programs. The Plan is a strategic overlay — coordinating and sequencing work across these instruments rather than replacing them — and instructs agencies to ensure sectoral regulators retain primary enforcement authority.

International Alignment

The Plan emphasises international engagement and alignment with global norms. Australia intends to participate in international fora on AI safety and norms, to join networks such as the International Network of AI Safety Institutes via the AISI, and to coordinate regional partnerships that reflect Australian values. The strategy references OECD definitions and international standards work, signalling a commitment to interoperability of approaches, cooperation on safety testing and joint efforts to address cross-border harms such as AI-enabled crime and misuse. The Plan therefore balances domestic, fit-for-purpose approaches with an intent to harmonise where possible with trusted international partners.

Implementation Timeline

Sources and References