**
As artificial intelligence (AI) continues its rapid evolution, an unexpected concern has emerged regarding the environmental impact of this technology: its significant water consumption. Analysts and environmentalists are raising alarms about the water footprint associated with AI operations, particularly as data centres—the backbone of AI processing—expand their influence and resource requirements.
The Thirst of Data Centres
Data centres, which host the servers powering AI applications such as ChatGPT and Google Gemini, require considerable amounts of water for cooling systems and electricity generation. Estimates regarding water usage vary widely, sparking a heated debate among industry experts. On one hand, Sam Altman, CEO of OpenAI, claims that a typical query from ChatGPT consumes less than 1/15 of a teaspoon of water. Conversely, research from the University of California in 2023 suggests that ChatGPT could use approximately 500ml of water for every 10 to 50 responses.
This discrepancy raises critical questions about the true scale of water consumption in AI operations. A report by the UK Government Digital Sustainability Alliance warns that if current trends continue, global water usage driven by AI could soar from 1.1 billion to 6.6 billion cubic metres by 2027. This staggering figure represents over half of the UK’s total water consumption, highlighting the potential strain AI could place on already scarce resources.
The Environmental Impact
As AI technology proliferates, concerns are mounting about its implications for local water supplies. Many data centres are situated in areas where water is already a limited resource, exacerbating the pressure on vulnerable ecosystems. A significant 68 percent of these facilities are located near protected biodiversity areas, which rely on clean water for their survival. The Government Digital Sustainability Alliance has expressed alarm, predicting that demand for fresh water could exceed supply by 40 percent by the decade’s end.
The environmental consequences of this water consumption are profound. Data centres, particularly larger operations, can consume as much water as entire towns. For instance, a mid-sized data centre may use the same amount of water as a small community, while a larger facility could require up to 5 million gallons daily—equivalent to the needs of a city with 50,000 residents. This reality poses serious challenges for local communities and ecosystems alike.
Debating the Crisis
While some experts warn of a looming crisis, others argue that fears surrounding AI’s water consumption may be exaggerated. Andy Masley, director of Effective Altruism DC, suggests that the water used for an individual AI interaction is negligible compared to everyday water needs. He emphasises that one would need to submit over 1,000 prompts in a single day just to marginally increase their water footprint.
However, other scholars, like Sam Gilbert from the University of Cambridge, caution against focusing solely on individual usage. Gilbert underscores the importance of examining the impact of data centres on local environments and water supplies. He advocates for greater transparency from companies regarding the environmental footprint of their operations.
Nick Couldry, a sociologist at the London School of Economics, contends that the increasing dependence on AI for everyday activities will inevitably lead to heightened demands on physical resources. He questions whether technology companies can sustainably manage this escalating consumption, given the addictive nature of their business models.
Seasonal Strain and Long-Term Solutions
The issue is further complicated by the seasonal nature of water availability. Shaolei Ren, an engineering professor at the University of California, notes that many data centres experience peak water usage during summer months, which can strain public water systems. Thames Water has already warned that these facilities could face restrictions during heatwaves, highlighting the urgent need for sustainable practices in water management.
Nevertheless, there is potential for AI to contribute positively to water conservation efforts. Innovative AI technologies are being employed to detect leaks and optimise water distribution systems. For example, a water company in Surrey has started using AI to minimise leaks across its network. The World Economic Forum has reported that AI-enabled solutions could lower water usage by up to 15 percent in the United States once fully implemented.
Google’s data centres, known for their efficiency, utilise air-cooling systems to reduce water requirements. A company spokesperson stated that Google is committed to replenishing 120 percent of the water it consumes on average, illustrating a proactive approach to sustainability.
Why it Matters
As the demand for AI technologies escalates, understanding the environmental implications, particularly regarding water consumption, is crucial. The balance between technological advancement and ecological sustainability is delicate. Stakeholders must prioritise transparency and innovation to ensure that the water needs of AI do not undermine the very resources upon which our communities and ecosystems depend. In an era where every drop counts, the industry’s response will shape the future of both technology and environmental stewardship.
