Briefing
Reference 15 · Primer · About 5 minutes
Download the MP3 · Leonard, an AI voice, reads a script written from the sources below.
Written from the National Archives, Finance and DTA pages, checked on 11 November 2026. Paraphrased, not reproduced. Check the current text before relying on it.
Before we record, every claim the hosts make is checked against the sources (this table was last updated 11 November 2026). Confirmed means a primary source or two independent reports; Reported means one party says so; Partly right, Disputed, Unverified and Not yet known mean what they say, and the hosts say so on air. This is the table as we left it.
The rules that apply to AI in Australia and for Australian organisations with European exposure, and the practical work each one calls for.
These are Agile Insights AI Strategy & Governance offerings. Which of them you need depends on where your organisation starts.
Sign up for our podcast at agile-insights.com.au.
Leonard Every government AI service still has to meet rules that never mention AI. Records law, data ethics, digital service standards and cloud policy all apply on day one. By the end of this episode, you will know what each of these quiet obligations asks of an AI project.
Leonard First, what they are. There are four. The Archives Act is a law, looked after by the National Archives of Australia. The Data Ethics Framework is guidance from the Department of Finance. The Digital Experience Policy comes from the Digital Transformation Agency, the D T A. It is mandatory, and it comes with four standards. The Whole of Government Cloud Computing Policy is also mandatory for non corporate Commonwealth entities.
Leonard The problem they solve is the gap between an AI pilot and a real service. A pilot can ignore records, accessibility and hosting. A live service cannot. Picture a chatbot that drafts letters to citizens. Is each draft a record? Can a person who cannot use the chatbot still get help? Where does it run, and how would you leave? None of those questions is about the model.
Leonard Who uses them? Records and information managers own the archives side. Data owners and privacy teams use the ethics framework. Service owners and designers answer to the digital experience standards. Architects and chief information officers make the cloud calls. Auditors and the archives regulator ask the questions afterwards.
Leonard So what is in them? Start with records. The Archives Act is technology neutral, and the National Archives says AI outputs are Commonwealth records. Drafts that only feed a final record can be short lived. Prompts should be kept where the business needs them, for example to show how a high value output was made. AI content accepted into a record without review should be flagged. And because AI tools are not recordkeeping systems, records must be captured into one that is. Business systems that include AI should go through the Archives' business system assessment framework.
Leonard Next, data ethics. The framework, published in December 2024, guides the ethical use of public data, including in new technologies. Training a model, retrieving documents and reusing administrative data are data ethics questions before they are AI questions. If your AI project runs on data collected for another purpose, start here.
Leonard Then digital experience. The policy took effect in January 2025 and was revised in December 2025. Its four standards cover service design, inclusion, access and performance, with dates that have already passed for most new services. The inclusion standard, for example, has applied to existing public facing services since January 2026. An AI front door is still a digital service, so all four apply. That includes a way in for people who cannot or will not use AI.
Leonard Finally, cloud. The cloud policy took effect on the first of July 2026. It makes cloud the default for new investment, and pushes agencies off legacy systems. It sets where most AI workloads will run. It does not settle vendor dependence or exit, so plan those yourself.
Leonard How does this connect to the rest of the stack? The D T A's AI Reference Architecture, from Reference 12, is where these land. Records sit in its traceability domain. The digital experience standards sit in its experience layer. Cloud sits in its platform layer, beside sustainability and vendor dependence. One tension to watch is retention. Security guidance may want chat content deleted, while records law may want it kept. Capture the record first.
Leonard Getting started. First, decide which AI outputs are records, and where they will be captured. Second, check each AI service against the four digital experience standards, including a non AI route. Third, write down where each AI workload runs and how you would move it. The primary sources are the National Archives, Finance and D T A websites, linked in the show notes.
Leonard Now, currency. As at November 2026, the cloud policy is the newest, in force since July. The archives guidance is updated from time to time. Check the current text before you rely on any detail here.
Leonard In short, AI changes the tool, not the obligations around it. If you'd like to know more, sign up for our podcast at agile-insights.com.au.
All of the reference series (16) · Follow AI Governance Reference in Apple Podcasts or any podcast app
Everything from Agile Insights: briefings, Making You Think and the reference series