AI and data centres are significantly increasing electricity demand and greenhouse gas emissions, making digital infrastructure a growing environmental concern. Technologies such as underwater data centres, liquid cooling, and waste heat recovery are improving energy efficiency and enabling more sustainable data centres. As digital infrastructure expands, managing electricity demand and emissions will be critical for long-term sustainability.
The rapid expansion of artificial intelligence (AI), cloud computing, and digital services is driving an unprecedented surge in data processing—and with it, a growing environmental footprint. Data centres, the backbone of the digital economy, are becoming increasingly energy-intensive, raising critical questions about sustainability, efficiency, and long-term resource use.
At the centre of this challenge is electricity demand. According to the International Energy Agency, data centres currently account for approximately 1–1.5% of global electricity demand, with AI-driven workloads expected to accelerate this growth significantly. Training and operating large-scale AI models requires vast computational power, often involving thousands of high-performance processors running continuously. As AI adoption expands across industries, electricity demand from digital infrastructure is projected to increase substantially.
This rising electricity demand translates directly into greenhouse gas emissions, particularly in regions where power generation remains carbon-intensive. Hyperscale data centres operated by major technology companies can consume as much electricity as small cities, making them a material contributor to global greenhouse gas emissions. Although many operators are investing in renewable energy, the pace of AI-driven demand continues to challenge decarbonisation efforts.
Water usage is emerging as an additional environmental pressure. Data centres rely on cooling systems to maintain stable operating conditions, with many facilities using evaporative cooling that consumes significant volumes of freshwater. In water-stressed regions, this creates competition for resources and expands the environmental footprint of digital infrastructure beyond emissions alone.
Technological innovation is playing a central role in advancing energy-efficient and sustainable data centres. Underwater data centres, such as those tested by Microsoft, use naturally low ocean temperatures to reduce cooling energy requirements. These systems have demonstrated improved energy efficiency and more stable operating environments.
At the same time, liquid cooling technologies are replacing traditional air-based systems, enabling more efficient heat management and supporting the higher computing density required for AI workloads. Edge and modular data centres are also gaining traction, allowing processing to occur closer to end-users, reducing transmission losses, and facilitating integration with renewable energy sources.
Waste heat recovery is another emerging solution. Instead of releasing excess heat into the atmosphere, some data centres are redirecting it to nearby buildings or district heating systems, improving overall energy efficiency and supporting more circular infrastructure models.
For businesses, these developments are increasingly relevant. Nearly all sectors—including finance, manufacturing, healthcare, and digital services—depend on data centres and AI systems. As reliance on digital infrastructure grows, companies are indirectly exposed to its environmental footprint, including electricity demand, greenhouse gas emissions, and resource use. This creates both operational and strategic implications, particularly as energy costs rise and expectations for sustainable digital operations increase.
Ultimately, AI and data centres are becoming a defining environmental challenge of the digital economy. While innovations such as underwater data centres, liquid cooling, and waste heat recovery offer clear pathways to improve efficiency, the scale of digital growth means that electricity demand and greenhouse gas emissions will remain a critical issue. The long-term sustainability of AI will depend not only on technological advancement, but on how efficiently and responsibly the infrastructure behind it is powered.
Read more at:
- International Energy Agency – Data Centres and Data Transmission Networks
- Microsoft – Project Natick (Underwater Data Centres)