Why this matters now
On 15 July 2026, Prime Minister Anthony Albanese announced the creation of a new Office of Artificial Intelligence, sitting inside his own department, the Department of Prime Minister and Cabinet. Its job is to coordinate a national framework (dubbed “Australian Standards”) covering AI’s impact on copyright, energy, productivity, education, and labour rights.
The pitch is that Australia will be the first country to bring these issues into a single national framework, rather than the patchwork approach most jurisdictions have taken so far.
For business leaders already using AI, or weighing it up, this is a signal worth pausing on. Not because the detail is settled (it isn’t) but because the direction of travel just became a lot clearer.
What was actually announced
A few things are worth separating out from the headlines.
The Office of AI will sit within PM&C and work with Industry Minister Tim Ayres and Assistant Minister for Technology Andrew Charlton to design the new standards and pull together AI-related work that’s currently scattered across government. It builds on existing work on a “digital duty of care” and on mitigating AI-enabled risks to children and national security, including the use of AI to spread disinformation.
What it doesn’t do, at least not yet, is settle the copyright question. That’s significant because Anthropic has said publicly that its $21.6 billion investment in Australian AI infrastructure depends on getting copyright clarity. The Prime Minister’s counter-argument is that clear guardrails will attract investment, not repel it. The problem though is that “clear” isn’t the same as “decided,” and businesses building on these tools are still operating in the gap between the two.
There’s also a live tension inside the government itself. Labor MP and former industry minister Ed Husic has publicly described reliance on big tech’s “social licence” as a path “doomed to failure,” arguing that voluntary self-regulation has already been tried, has already failed once, and won’t hold financially for firms with every incentive to move fast. That’s not a fringe view, it’s coming from someone who used to hold the industry portfolio.
What it means if you’re the one accountable for AI in your business
None of this changes what’s actually required of you today. But it does tell you three things about tomorrow.
Regulation is coming, and it’s coming as a system, not a checklist. A single national framework, rather than a series of unrelated rules, means organisations that have been treating AI governance as “add a policy when someone asks” are going to find that approach doesn’t map onto what’s coming. Governance built as a structured management system (the kind ISO 42001 already gives you a template for) will translate far more easily into whatever “Australian Standards” ends up meaning in practice.
Copyright and IP risk in AI use isn’t going away as a live issue. If a $21.6 billion infrastructure investment is contingent on this being resolved, it’s not a minor detail, it’s a live risk category. If your organisation is using generative AI in content, product, or customer-facing work, this is a governance question you should already be able to answer, not one to wait on Canberra for.
“We’ll self-regulate until the government makes us stop” is no longer a safe assumption. With senior figures inside the governing party openly saying self-regulation won’t hold, the sensible bet for any board or executive team is that mandatory standards land eventually. Organisations with governance already in place won’t be scrambling when they do. Organisations without it will be building under pressure, on someone else’s timeline.
Strong governance should make better decisions easier, not slower — and that’s just as true when the regulatory ground is still moving.
Practical steps to take now, while the framework is still being written
Get your AI inventory in order
You can’t assess exposure to a national framework you haven’t mapped your own AI use against. Know what tools, vendors, and use cases exist across the business today.
Separate the “settled” risks from the “pending” ones.
Privacy obligations are settled law. Copyright and IP use in AI training and outputs is not. Treat them differently, and don’t let uncertainty on one become an excuse to ignore the other.
Assign real ownership.
A national Office of AI is a coordination body, not a compliance function for your organisation. Someone inside your business needs to own AI governance outcomes. This means having someone to oversee approvals, monitoring, incident response, regardless of what Canberra does next.
Build to a standard, not just a policy.
ISO 42001 gives you a structured, internationally recognised way to demonstrate governance maturity now, and a natural on-ramp if and when domestic standards become mandatory.
Watch this space, but don’t wait on it.
The framework will take time to move from announcement to detail to law. Good governance doesn’t need to wait for that timeline to be useful.
Where Governance Works can help
Governance Works specialises in AI governance, ISO 42001 implementation, ISO 27001 support, IT security advice, and fractional CISO and CIO services. I help organisations build governance that’s right-fit for their size and risk profile, practical enough to implement now, credible enough to hold up when the regulatory picture firms up.
If your organisation is trying to work out what today’s announcement actually means for you, the most useful next step is usually a focused conversation about where you stand today.
Book a consultation to talk through your AI governance priorities.
An audit is a quick first step. If you don’t know what you have, you can’t put governance in place.