Approvals Aren’t Enough
The Victorian government recently promised data centres less regulation to woo them over from NSW, where the majority of data centres currently sit. Call it some friendly competition between states; the federal government is also canvassing tying fast-track approvals with sustainability and renewable energy benchmarks to qualify.
And it’s true - there’s a backlog of data centre plannings approvals. There’s so much money piling into the space, how could there not be? Australia has the second largest pipeline of projects in the world. When a bucket of gold drops in your lap, everyone scrambles to stop it leaving for more investor-friendly laps.
The problem is that clean infrastructure is not solely an approvals problem. It is a coordination and partnership problem, and treating it like a simple planning bottleneck risks locking in the very outcomes policymakers say they want to avoid.
Approvals answer whether a project proceeds (or not).They do not answer the harder, more important ones: where should it go, what should it run on, and how does it interact with the energy system over the next 20 years?
Why data centres break the traditional planning model
Data centres sit awkwardly inside Australia’s planning frameworks because they don’t behave like the infrastructure we’re used to approving.
They draw industrial-scale electricity loads, require continuous cooling, and rely on firm power. They are also unusually mobile: developers can choose between regions, states, and countries based on power, fibre, water, policy and politics.
Planning systems largely treat data centres as oversized warehouses or office buildings (which, to be fair, they also are), but without considering their flexibility.
That mismatch matters. A 100-megawatt data centre is not just a land-use decision, but a grid decision, a water decision, a regional development decision and, increasingly, a climate decision. Once built, these facilities exert a gravitational pull on electricity markets and infrastructure investment for decades.
Approving them in isolation assumes the system will sort itself out later, which is more or less the attitude of the Victorian government.
The limits of “approve it and figure it out”
Left to their own devices, data centres will do what rational businesses do: build where power is cheapest and most reliable today, not where the cleanest energy might exist tomorrow. Market mechanisms will certainly help solve water and power efficiencies, but siting is far less certain.
That dynamic is already visible internationally. Where compute demand arrives faster than planning and grid upgrades, electricity prices rise, fossil generation runs harder, and public concern escalates. In those conditions, even well-intentioned renewable projects become politically fragile. Witness Singapore and Ireland putting moratoriums on new builds in recent years while policymakers caught up.
Fast approvals without coordination don’t accelerate the transition. They produce:
digital load clustering on constrained, carbon-intensive grids
existing renewable energy still stranded behind transmission bottlenecks
rising costs passed on to consumers
communities that feel infrastructure is imposed rather than designed with them
Once those dynamics take hold, political support erodes quickly. See: NSW government banning new builds in Macquarie park, and launching an inquiry last month.
Better Design = Less Headache
Approvals are necessary but insufficient. There’s a few actions we can take to avoid expensive remediation down the track.
1. Energy partnerships
Large data centres should be deliberately linked to firmed renewable supply — through long-term PPAs, co-located storage, or direct connections to renewable generation. They need to continue underwriting new clean capacity and reducing grid stress but connecting as close to the source as possible.
2. Planning partnerships
Digital infrastructure should be visible in energy planning, and energy constraints should shape digital siting decisions. Shared spatial planning between energy market bodies, state planners and digital infrastructure agencies would prevent today’s ghost grids and digital islands. In fact, these tools already exist!
3. Policy partnerships
A national framework that sets expectations upfront — on disclosure, siting, energy sourcing, incentives and regional contribution — gives industry certainty and communities confidence. Without it, every project becomes a one-off negotiation fought in the media and the courts. We hope the government’s National AI plan doesn’t merely nod at sustainable data centres, but leverages them for our energy transition.
Now, Now, Now
This is such a pivotal stage. Data centres take 18-24 months to build and operate for decades. The approvals being issued today will shape Australia’s energy mix, regional economies and emissions profile well into the 2040s.
The federal government has flagged national data-centre principles as part of its broader AI agenda. That creates a narrow but critical window to move beyond speed and toward structure.
If we get this right, data centres can become anchor customers for renewables, catalysts for regional development, and partners in managing a more decentralised grid. If we get it wrong, they could be blamed — fairly or not — for higher prices, grid congestion and stalled climate goals.
