Australia’s AI Policy Pivot
For years, Australia’s approach to AI came with a 'hands-off' posture. Last week, however, there was a significant change. Prime Minister Anthony Albanese’s ‘AI in Australia’s Interests’ address at the University of Sydney, signalled a pivot toward a mandatory framework. This move seeks to ensure Australia is not reduced to a data warehouse for offshore AI products, and instead develops sovereign capability, trusted infrastructure and governance needed to capture the benefits of AI on Australian terms.
We Need Sovereignty
Synlaw has long argued, the priority is not merely speed, but sovereignty, security, and compliance. There’s a sense of urgency in this shift, and it comes from Australia's current lack of digital sovereignty.
In fact, Australia recently lost access to advanced AI models because they were revoked by the Trump administration. Relying on offshore-owned frontier models means Australia’s productivity could be at riskat any moment.
As the Prime Minister rightly pointed out, Australia cannot afford to be ‘subcontracting our sovereignty and security to the control of foreign monopolies’.
The Office of AI and National Standards
In the first steps to address these risks, the Government has established the Office of AI within the Department of the Prime Minister and Cabinet, and will develop national standards. These standards will initially target large data centres, imposing obligations to underwrite their own power supply, remain water-efficient, and pay their share of connection costs.
Whilst the AI vision for Australia scores high on ambition, Lucy Poole, Deputy CEO of the Digital Transformation Agency, recently highlighted the risk of leaving a ‘gap between having a framework and having evidence that the framework is working’. The follow-through on the standards will be critical.
Consumer Priorities
The government has also outlined new AI consumer safety priorities.
These priorities include legislating a Digital Duty of Care, which puts the onus on AI companies to build in ‘safety by design’. This agenda also prioritises:
Privacy: A second tranche of reform to modernise personal data protection laws.
Automated Decision-Making (ADM): Developing a framework to ensure fair and transparent government decisions - an area where Synlaw has consistently advocated for human accountability.
These priorities align with Synlaw’s stance that compliance must be built-in, not bolted on. A framework for ADM is critical to ensure that ‘human in the loop’ protections are present from the outset, rather than treated as a retrospective audit task or operational afterthought.
Judgment as the Ultimate Asset
Many important questions in this space remain unanswered, including how we will avoid administrative bottlenecks.
Our view is that AI will always produce a bottleneck unless integrated appropriately. Without properly considered workflows and a clear path to a compliant outcome, AI merely fuels silos with content at speed.
Importantly, the Prime Minister stated that his goal is to provide ‘flexibility to keep pace with change’ without the ‘bureaucracy having to contemplate every modification’. He promised the new AI Standards would ‘speed up approvals’ and ‘streamline the process for verifying compliance’.
However, the risk of bottlenecks is still real. AI does not solve productivity on its own. We believe that judgment is the asset, and AI is the accelerator. Whilst the Prime Minister’s speech is heavily focused on standards and approvals, it’s important to be aware that a framework is only as good as the systems that enable it.
The Prime Minister is correct that we cannot sit back and allow AI to ‘take its course’. As we move toward the legislated standards in 2027, there must be close attention on whether organisations have the ‘control tower’ required to automate while maintaining judgment and compliance.
The ALP conference this week may provide some early indications of how the Prime Minister’s vision translates to policy, legislation and institutional design – and we will be watching closely.