New AI Technical Standard Released to Guide Responsible Government Use

Developed in close consultation with government agencies and informed by international best practice, the Australian Government AI Technical Standard complements the DTA’s broader suite of AI guidance available to the Australian Public Service (APS). It aims to embed responsible AI principles while helping agencies harness the benefits of AI to improve service delivery and public trust.

“Our goal is to position Australia as a global leader in the safe adoption of AI without stifling innovation,” said Lucy Poole, General Manager of Digital Strategy, Policy and Performance at the DTA. “This standard offers a flexible, lifecycle-based approach that supports agencies to integrate AI responsibly within their existing frameworks.”

What the Standard Covers?

The standard outlines technical requirements and recommendations across the full lifecycle of AI systems—from conception and deployment to retirement. It applies to all delivery models, including in-house development, vendor-supplied solutions, pre-trained AI models, and managed services.

The framework is divided into three core phases:

• Discover – Designing and preparing AI systems. This includes defining purpose, managing ethical risks, ensuring data quality and privacy, selecting and training models, and performing robust evaluations.

• Operate – Implementing AI systems. This phase covers integration into business systems, secure operation, continuous monitoring, and responsiveness to performance issues or emerging risks.

• Retire – Managing the responsible decommissioning of AI systems. This ensures compliance with data retention policies, risk assessments, and proper transitions where necessary.

Throughout all phases, the standard emphasises the importance of public trust, human oversight, transparency, and inclusive design.

“This isn’t about creating extra red tape,” added Ms Poole. “The standard is designed to work with what agencies are already doing, allowing responsible AI practices to be embedded into current governance and delivery models.”

Next Steps for Agencies

Agencies are encouraged to begin incorporating the standard into the design and management of both existing and future AI systems. The DTA will continue supporting agencies as they embed the standard within their operational and assurance practices.

Explore the full standard and supporting materials:
🔗 Technical Standard for Government’s Use of Artificial Intelligence – digital.gov.au

An overview is also available via the Australian Government Architecture:
🔗 AI Technical Standard – AGA