25 September 2026
4 min read
#New South Wales Government, #Government, #Governance, #Data & Privacy
Published by:
Artificial intelligence (AI) has been around for some time and is becoming a practical feature of everyday public sector service delivery. Across the NSW Government, agencies are increasingly using AI to improve productivity, support decision-making and deliver more targeted services. While that growth brings clear opportunities, it also increases the need for disciplined governance, transparent records and strong assurance processes.
Just over one year ago, the NSW Government established the Office of Artificial Intelligence (Office of AI) to coordinate AI initiatives across government and strengthen whole-of-government AI leadership. The Office of AI’s role is significant because it reflects a move from experimentation to more deliberate, coordinated and accountable use of AI across NSW.
In 2025, the NSW Auditor-General identified 357 AI tools in use across 21 of the 26 major NSW public sector agencies reviewed. These figures demonstrate that AI is already embedded in many government operating environments, even where formal agency-level strategy and governance are still maturing.
For agencies, AI adoption is no longer just a technology issue. It also raises legal, governance, procurement, privacy, cyber security and workforce capability considerations. Agencies should therefore include AI as part of their core risk and assurance framework, rather than treating it as a standalone digital initiative.
The NSW Auditor-General identified several challenges and barriers to responsible AI adoption, including:
These risks highlight why AI strategy needs to be practical and operational. A policy that sits outside day-to-day decision-making will not be enough. Agencies need clear ownership, an up-to-date inventory of AI use cases, risk-based assessment processes, procurement controls, staff guidance and regular reporting to senior leaders.
1. Check whether AI meets the mandatory requirements in the policy
Agencies should check what AI tools are being used, by whom, for what purpose and with what limitations. This should include both approved enterprise tools and lower-profile uses, such as embedded AI features in existing software. Agencies should also ensure that all AI use cases are registered in their AI Assessment Framework platform and that they maintain accurate and complete records through the AI system lifecycle.
If an agency has not reviewed its AI governance since the release of the current policy settings, now is an appropriate time to do so.
2. Embed AI into existing risk, procurement and cyber processes
AI governance is more effective if it is integrated into processes agencies already use. Procurement teams should understand when AI is present in a product or service, how vendor AI claims will be tested, what contractual protections are needed and how data will be handled. Cyber security and privacy teams should be involved early, particularly where AI tools use sensitive information, automated recommendations or external platforms.
This approach can also reduce duplication. Rather than creating a separate AI governance process for every issue, agencies can build AI-specific questions, thresholds and approvals into existing project, procurement, privacy, data and cyber security frameworks. This makes governance easier to apply and more likely to be followed in practice.
3. Build AI literacy and senior oversight
Senior leaders need visibility to understand where AI is being used, what benefits are expected and what risks require active management. Regular reporting to senior management or a governance committee can help ensure AI initiatives remain aligned with agency objectives, legal obligations and public expectations.
AI literacy is equally important. Staff need practical guidance on what AI tools can and cannot be used for, how to check AI-generated outputs, when human review is required and when advice should be sought. This is particularly important where AI outputs may influence decisions affecting members of the public.
Ultimately, AI will continue to transform the NSW public sector. Agencies that invest early in clear governance, transparent records, staff capability and risk-based assurance will be better placed to realise the benefits of AI while maintaining public trust. We have been assisting government agencies to manage these issues and adopt AI with appropriate safeguards and frameworks in place.
If you have any questions about AI use in government, please get in touch with us.
Disclaimer
The information in this article is of a general nature and is not intended to address the circumstances of any particular individual or entity. Although we endeavour to provide accurate and timely information, we do not guarantee that the information in this article is accurate at the date it is received or that it will continue to be accurate in the future.
Published by:
