If you build or sell an AI product in Australia you already know the awkward part of the regulatory map: no single regulator owns it. A recommender system is a consumer-protection question for the ACCC, a privacy question for the OAIC, an online-safety question for eSafety and a media question for the ACMA, often at once. DP-REG is the body that exists to stop those four talking past each other. It has been going since March 2022 and, until this year, most of what it published was about AI.
What it actually did in June
Two documents, two days apart. The MoU is the governance one. The ACCC's announcement
says it formalises how members will share information and work together on cross-cutting
issues, including scams, privacy, online safety and competition concerns
, and is careful
to add that While each regulator will continue to carry out its statutory functions
independently
the agreement supports a more structured approach to cooperation. The
signed document defines its own subject broadly, covering internet search engines,
digital content aggregators, social media services
and more.
The MoU also names the problems it is for. Members, it says, face common challenges
including addressing scams and emerging consumer harms, balancing innovation with
appropriate protections, and responding to the market power of large, multinational digital
platform operators
. That is a fair description of the AI policy debate. It does not use
the word.
The priorities document is the more interesting of the two, because it carries dates.
DP-REG set itself four measurable commitments, and the first is already met: Execute a
memorandum of understanding between DP-REG member regulators by 30 June 2026.
The
signature pages of the MoU we downloaded run from 28 May to 2 June 2026, so all four
regulators had signed nearly a month before the deadline, and the announcement followed on
17 June. A regulator meeting its own published deadline early is not a story anyone usually
writes, so it is worth saying plainly that this one did.
The three commitments still on the clock
The other three metrics are all worded for the current year, which makes them checkable by the end of it:
- a DP-REG civil society engagement activity, with attendee satisfaction measured by a post-event survey;
- a series of learning and development sessions for member regulators;
Release of a paper regarding dispute resolution this year.
The dispute-resolution paper is the one to watch if you are building anything consumer facing. Where a user complains about an automated decision, whose complaints process they land in is currently a function of which regulator's statute the harm happens to fit. A joint paper on that is the sort of thing that changes how a product team designs an appeals flow.
What changed, and being careful about what it means
The absence stands out because of what came before it. DP-REG's own news page lists its earlier output, and the last three substantive items are AI:
| Published | DP-REG output |
|---|---|
| 11 September 2023 | Joint statement on AI |
| 23 November 2023 | Working papers on algorithms and AI |
| 19 September 2024 | Working paper on multimodal foundation models |
| 15 June 2026 | 2026-28 strategic priorities. No AI mention |
| 17 June 2026 | Memorandum of Understanding. No AI mention |
Our view, labelled as such, and it is narrower than the headline invites.
This is not evidence that DP-REG has stopped caring about AI, and we are not going to write
it that way. Read the two documents and the reason for the absence is visible in their
shape: they are framed by capability rather than by technology. The priorities
commit members to increase regulatory capability, improve information sharing, collaborate
on regulatory development, and collaborate to support industry innovation that leads to
productivity and efficiency benefits while ensuring that implementation protects
Australians
. Those are containers. AI fits inside them without being named, and so does
everything else.
What is true, and worth a builder's attention, is narrower and harder to argue with: there is no longer an AI-specific commitment with a date on it. Between 2023 and 2024 there was published joint AI analysis you could read and cite. For 2026-28 there is a coordination framework, and whether AI gets joint attention inside it is now a matter of what the members choose to prioritise rather than something the forum has written down.
That distinction matters more than it sounds for anyone trying to anticipate Australian AI regulation. A named workstream is a signal you can plan against. A capability framework is not, however sensible it is as governance.
What this does not tell you
DP-REG publishes sparingly and its members regulate AI constantly on their own account: eSafety's industry codes and standards, the OAIC's guidance on facial recognition, the ACCC's digital platform services work. Nothing here says any of that has slowed, and this masthead has covered several of those threads separately. The absence we are reporting is specifically in the forum's two June 2026 documents, which is a claim about text, not about intent. We have not asked DP-REG or any member regulator why AI is not named, and their answer might be the obvious one: that a governance instrument does not enumerate subject matter.
How we did this
We read the DP-REG 2026-28 strategic priorities page, the forum's Memorandum of Understanding page, the ACCC's announcement, and the signed MoU PDF, which we downloaded and converted to text: 2,332 words. The absence claim was tested against all four by case-insensitive search for artificial intelligence, AI as a standalone word, algorithm, machine learning and generative, and returned zero across every one. Testing an absence against a summary rather than the signed document would not have been worth publishing, which is why the PDF was fetched.
The publication dates in the table are the dates DP-REG's own pages carry. The MoU signature dates are read from the signature pages of the PDF, not from the announcement. The observation that the priorities are framed by capability rather than technology is ours, and so is the reading of what that means for anyone planning against Australian AI regulation.
One access note for anyone following this trail: the ACCC's news centre is a JavaScript-rendered listing with topic facets and cannot be read with a plain fetch, and `accc.gov.au/sitemap.xml` returns 404 with no sitemap declared in robots.txt. The forum also has its own site at dp-reg.gov.au, which is where the primary documents are.
Sources
- Digital Platform Regulators Forum, DP-REG's 2026-28 strategic priorities (last updated 15 June 2026, read 28 August 2026): the retained 2024-26 goals; the capability, coherence and emerging-risk priorities including the right-size regulation wording; and all four metrics, including the 30 June 2026 MoU deadline and the dispute-resolution paper.
- Digital Platform Regulators Forum, DP-REG Memorandum of Understanding, signed by all members (PDF, downloaded and read 28 August 2026, 2,332 words): the four member regulators; the definition of a digital platform; the common challenges wording; and the signature dates of 28 May to 2 June 2026.
- ACCC, Joint Statement: Regulators strengthen joint oversight of digital platforms (17 June 2026, read 28 August 2026): that the MoU formalises information sharing on scams, privacy, online safety and competition; that members continue to exercise their statutory functions independently; and the background on DP-REG's March 2022 establishment.
- Digital Platform Regulators Forum, News and media (read 28 August 2026): the forum's full publication list and the dates of the joint statement on AI (11 September 2023), the working papers on algorithms and AI (23 November 2023) and the working paper on multimodal foundation models (19 September 2024).
Spotted an error? Tell us and we will check it against the documents above and log the outcome.