National and state AI requirements

On this page

AI is a focus area of governments, globally. There are specific laws, policy and guidance from the Australian and NSW governments that must inform our approach to adopting AI technologies across NSW Health.

Technology is evolving rapidly, which means government approaches are iterative, and the regulatory and policy landscape is continually shifting.

NSW Health must comply with all applicable laws when developing and using an AI systems. Relevant legislation is noted throughout this AI Framework, and a useful list of key legislation that applies to AI in NSW is available from Digital NSW.

Below, are the existing NSW and national AI reference documents, that should be considered and applied in NSW Health.

NSW Government AI laws, policy and guidance

  • Circular: Use of Artificial Intelligence by NSW Government Agencies
    This NSW Department of Customer Service Circular (DCS-2024-04) stipulates the mandatory requirements for all NSW government agencies using AI. Compliance with the NSW Government Artificial Intelligence Ethics Policy and NSW Government Artificial Intelligence Assessment Framework is mandatory for all agencies using AI, including NSW Health.
  • Artificial Intelligence Strategy
    This strategy directs the development and use of AI by NSW government agencies. It aims to support agencies to innovate with AI, while making sure it is used safely and responsibly.
  • Artificial Intelligence Ethics Policy
    This policy outlines 5 key principles that are mandatory for NSW government agencies to apply best practice when using AI: community benefit, fairness, transparency, accountability, privacy and security. NSW government agencies that are using AI cannot breach these ethical principles.
  • Artificial Intelligence Procurement Essentials
    Provides NSW government buying teams with key considerations for sourcing and managing AI systems.
  • Digital Assurance Framework
    This framework aims to ensure the NSW Government's digital projects are delivered on time and on budget through the implementation of a risk-based independent assurance framework. Digital assurance is mandatory for digital projects (including projects using AI) with budgets of $5 million or more (regardless of funding source) and for projects funded by the Digital Restart Fund. Digital assurance is recommended for all projects of strategic importance or that are high risk (as determined by the risk profile criteria set out in the Digital Assurance Framework).
      • Artificial Intelligence Assessment Framework
        A risk self-assessment framework for NSW government agencies to use to ensure responsible design, development, deployment, procurement and use of AI systems. It has been mandatory for all NSW government agencies to use when designing, developing, deploying, procuring and using AI systems since March 2022.

      The framework enables agencies to evaluate the risk associated with their AI systems and understand the mitigation strategies needed. It also supports agencies to determine when they need to present their solution to the NSW Artificial Intelligence Review Committee for further review.

      All NSW government agencies must submit an annual Artificial Intelligence Assurance Compliance Plan to the Department of Customer Service. This outlines the progress of each agency in ensuring their governance and assurance functions comply with the mandate to assess AI system risks using the AI Assessment Framework.

      Guidance to support the assessment framework:

      National AI guidance

      Australia’s Artificial Intelligence Ethics Principles
      These principles create a broader national framework for all sectors (public and private) and are designed to ensure AI is safe, secure and reliable. The 8 AI ethics principles apply to NSW Health activities, which help:

      • achieve safer, more reliable and fairer outcomes for all Australians
      • reduce the risk of negative impact on those affected by AI systems
      • businesses and governments to practice the highest ethical standards when designing, developing and implementing AI.

      National Framework for the Assurance of Artificial Intelligence in Government
      This framework provides a nationally consistent approach for the assurance of AI use in government. It is a joint publication of the Council of Australian Governments, including the Australian Government and each State and Territory. This document is advisory rather than prescriptive.

      It stipulates that to effectively apply the ethical principles, governments should consider the following cornerstones for their assurance practices:

      • Governance
      • Data governance
      • A risk-based approach
      • Standards
      • Procurement

      Voluntary AI Safety Standard

      This standard was prepared to align with international standards and consists of 10 voluntary guardrails to support human-centred AI deployment. It has largely been replaced by the below 'Guidance for AI Adoption: Implementation Practices'.

      Guidance for AI Adoption: Implementation Practices

      The Department of Industry, Science and Resources provides 6 essential practices for responsible AI governance and adoption:

      • Decide who is accountable: establish end-to-end accountability and robust AI governance.
      • Understand impacts and plan accordingly: ensure stakeholder rights and fair treatment.
      • Measure and manage risks: implement AI-specific risk management.
      • Share essential information: Ensure appropriate transparency and explainability.
      • Test and monitor.
      • Maintain human control: integrate meaningful human oversight and control of AI.
      Back to top