AI's Hidden Cost: Power, Water, Jobs at Risk

The rise of Artificial Intelligence is intrinsically linked to the expansion of data centers, which are crucial for AI 'inferencing' and 'training'. These facilities are rapidly increasing their consumption of electricity and water, prompting comparisons to the "satanic mills" of the first Industrial Revolution due to their resource appetite and potential impact on employment.

Globally, data centers consumed 1.5% of all electricity in 2024, a figure expected to grow significantly. In Australia, data center energy usage on the east coast was 2% in 2025 and is projected to triple by 2030. This escalating demand is putting pressure on energy grids, leading to concerns about grid stability and the origin of power sources.

Environmental worries extend to water usage, particularly in drought-prone regions. While some facilities are adopting water-saving technologies and recycled water, the overall consumption remains a concern for local communities. Public opposition to new data centers is growing, with some regions enacting moratoriums on new developments.

Technological innovations and regulatory efforts are underway to address these challenges. Solutions include locating data centers in remote areas powered by renewables, utilizing underwater facilities, or exploring entirely new computing technologies like biological-hybrid chips. Governments are also exploring regulations to define data centers' social license to operate.