Australia's enthusiastic push to secure a leading role in the global artificial intelligence revolution is facing a formidable and growing obstacle: the immense environmental footprint of the technology itself. The datacentres required to develop and run advanced AI systems are consuming staggering amounts of electricity and water, raising serious concerns about the nation's climate commitments and resource security.
The Insatiable Appetite of AI Infrastructure
While the promise of AI drives investment and political will, the physical infrastructure needed to support it operates with a voracious appetite for resources. Climate and environment reporter Petra Stock, in conversation with Nour Haydar for the Guardian's Full Story podcast, has highlighted the escalating worries among experts and communities. These facilities, housing thousands of high-performance servers, do not just require vast quantities of power to operate but also significant volumes of water for cooling systems to prevent overheating.
The central question now being asked is whether this exploding demand could fundamentally undermine Australia's pledged transition to a net zero emissions economy. Concurrently, the strain on local drinking water supplies, particularly in regions already facing scarcity, presents a critical challenge to sustainable development.
Power Drain vs. Climate Pledges
The electricity required to train and operate large AI models is monumental, often compared to the consumption of small cities. This surge in demand from a rapidly expanding sector threatens to overload grids still reliant on fossil fuels and complicate the planned phase-out of coal and gas plants. If not managed with a deliberate shift to renewable sources, the AI boom could lead to increased greenhouse gas emissions, putting the country's net zero ambitions in serious jeopardy.
Water: The Overlooked Resource Crisis
Perhaps less widely recognised is the profound thirst of massive datacentres. Cooling towers evaporate millions of litres of water annually, drawing directly from municipal supplies or local water tables. In a continent prone to drought, this consumption places acute pressure on drinking water reserves, affecting both communities and ecosystems. The issue transforms the narrative of AI from a purely digital endeavour into one with tangible, physical consequences for Australia's environment.
The trajectory is clear: without stringent regulation, investment in green infrastructure, and technological innovation in cooling efficiency, the cost of Australia's AI ambition may be paid for in strained power networks and depleted water reservoirs. The conversation started by Stock and Haydar underscores a vital need for a sustainable roadmap that aligns technological progress with environmental responsibility.
