AI’s Hidden Cost: How ChatGPT & Datacentres Impact the Environment & Australia’s Resources

The surge in demand for artificial intelligence is creating a parallel, and rapidly escalating, demand for power. Globally, datacentre energy consumption is growing at four times the rate of all other sectors, and is projected to exceed Japan’s entire electricity usage by 2030, according to the International Energy Agency report. This isn’t a future problem. in Australia, energy market operators anticipate datacentre energy demand will triple within five years, surpassing the electricity used by the nation’s entire fleet of electric vehicles by the end of the decade.

The environmental cost of this growth is becoming increasingly apparent, prompting questions about the sustainability of the AI boom and even calls for consumers to “opt out” of using AI tools altogether, as seen with the QuitGPT movement. Beyond electricity, authorities are also forecasting significant strain on drinking water supplies due to the cooling requirements of these massive facilities.

The Scale of the Energy Demand

Generative AI models – those capable of creating text, images, and video – are particularly energy intensive. Studies suggest they consume “orders of magnitude” more energy than traditional computing methods, with estimates ranging from five times more to significantly higher depending on the specific model and the complexity of the query according to a report to the Australian Parliament. This isn’t simply about running the finished AI product; the training of these models is a particularly energy-intensive process.

Ketan Joshi, a climate analyst associated with the Australia Institute, highlights the inefficiency of consumer software that generates content. He argues that using AI chatbots to answer questions consumes far more energy than a simple web search or calculation, comparing it to choosing an SUV over a bicycle for a short trip. “You might still get the shopping done, and that single trip alone may not even look all that bad in terms of cost or emissions, but what happens when that’s all of your trips, and when all of society starts doing this?” he asks.

The numbers are substantial. A study published in the journal Patterns estimates AI’s global carbon footprint at 32.6 to 79.7 million tonnes of CO2 emissions in 2025, with water usage reaching 312.5 to 764.6 billion litres – comparable to global bottled water consumption as detailed in the study.

Impact on Australia’s Energy Transition

In Australia, the growth of datacentres is projected to hinder the country’s progress towards net zero emissions and potentially increase power costs for consumers. The Clean Energy Council, Electrical Trades Union, Australian Conservation Foundation (ACF), and Climate Energy Finance have jointly proposed “public interest principles for datacentres” that emphasize investment in renewable energy sources and responsible water usage. Adam Bandt, the ACF chief executive, argues that tech companies should be compelled to invest in the infrastructure needed to support their operations, rather than placing the burden on consumers and the environment.

The physical manifestation of this demand is the rapid proliferation of datacentres – large, warehouse-like facilities requiring 24/7 operation and substantial cooling systems. These facilities are often clustered together, raising concerns about their impact on local communities, and wildlife. Dr. Bronwyn Cumbo, a researcher at the University of Technology Sydney, notes that communities are increasingly seeking greater engagement and transparency from datacentre operators.

Beyond Energy: Water and Local Impacts

The environmental impact extends beyond electricity consumption. Datacentres require significant amounts of water for cooling, placing additional strain on already stressed water resources, particularly in arid regions. The Guardian reported in December 2025 on the growing concerns surrounding water usage by datacentres in Australia.

Local communities are also facing challenges related to noise pollution from continuous air conditioning systems and the aesthetic impact of these large-scale facilities. Some communities have actively campaigned against the construction of new datacentres in their areas, highlighting the necessitate for greater community involvement in the planning process.

The Challenge of Opting Out

While the QuitGPT movement represents a direct protest against the ethical concerns surrounding AI, the practicalities of “opting out” are complex. AI tools are becoming increasingly integrated into everyday software, from workplace applications to chatbots used by banks and government agencies. Generative AI is even appearing in supermarket self-checkouts and facial recognition systems.

Prof Jeannie Paterson, co-director of the Centre for AI and Digital Ethics at the University of Melbourne, acknowledges the difficulty of avoiding AI altogether, but emphasizes the importance of expressing public opinion about its development and deployment. “We still have a chance to express our views about what and how we want AI to be used.”

Individuals can take small steps to reduce their reliance on AI, such as unsubscribing from AI platforms, excluding AI results from search queries, and avoiding unnecessary or energy-intensive tasks. However, the pervasive nature of AI makes complete avoidance increasingly challenging.

A Question of Collective Action

Ketan Joshi argues that opting out can be a “meaningful act of resistance,” not just in terms of reducing energy demand but also in fostering a broader collective action against what he describes as a “corrosive, harmful industry.” He cautions against simply switching between AI platforms, as exemplified by QuitGPT’s promotion of Anthropic’s Claude after encouraging users to cancel ChatGPT, viewing it as a potentially cynical tactic that doesn’t address the underlying issues.

The conversation surrounding AI and its environmental impact is evolving, with growing recognition of the need for greater transparency from tech companies and more robust regulation. Raising awareness and encouraging critical thinking are crucial steps towards ensuring that the benefits of AI are not achieved at the expense of the planet and local communities. The next steps will likely involve increased scrutiny of datacentre energy and water usage, as well as ongoing debate about the ethical and environmental implications of this rapidly developing technology.