Google promises to return more water than it consumes amid AI data centre backlash

As artificial intelligence fuels an unprecedented data centre boom, Google is stepping up efforts to address one of the industry's most contentious challenges: water consumption.

The technology giant has announced a new set of commitments aimed at reducing the impact of its US data centres on local water resources, including a pledge to replenish more water than its facilities consume by 2030. The move comes amid growing criticism of AI infrastructure projects, which require vast amounts of electricity and water to power and cool the hardware behind advanced AI systems.

Google argues that the overall contribution of its data centres to US water consumption remains relatively small. However, the company acknowledged increasing public concern and said it is taking additional steps to protect local water resources and improve transparency around its operations.

Google's five new commitments for water use at its data centres

In a blog post outlining its latest strategy, Google detailed five commitments that it says will strengthen water stewardship across its data centre operations.

More from Tech

The most significant pledge is a commitment to replenish more water than the company consumes at its data centre sites by 2030. According to Google, it replenished more than seven billion gallons of water in 2025 through various conservation and restoration projects. The company currently supports 165 water stewardship projects across 97 watersheds in the United States.

Google said these projects are expected to replenish more than 19 billion gallons of water annually by the end of the decade, more than double the amount of water the company consumed in 2024. In a separate paper, the company stated that it aims to replenish 120 per cent of the water used at its data centre sites by 2030.

The second commitment focuses on upgrading local water and wastewater infrastructure. Google said it has committed more than $500 million towards water supply, wastewater treatment and water reuse projects in communities where it operates or is building data centres. These investments include projects designed to improve local water availability and reduce losses from ageing infrastructure.

Third, Google pledged to protect vulnerable watersheds by prioritising air-cooled technologies in areas where water resources are under pressure. The company said it uses a data-driven framework to assess watershed health before building new facilities and will only consider water-based cooling systems where local resources are considered resilient. In higher-risk regions, it plans to rely on air cooling or recycled water instead.

Google also promised greater transparency around water use. The company noted that it was the first major cloud provider to publicly disclose annual water consumption data for its data centre locations and said it will continue publishing such information.

Finally, the company said it would accelerate efforts to use alternative water sources, including reclaimed wastewater. One example cited was a partnership in Douglas County, Georgia, where treated wastewater is being reused to cool data centre facilities.

Alongside these measures, Google announced an additional $17 million investment in new water stewardship projects across Georgia, Iowa, Michigan, Minnesota, Missouri, Nebraska and Texas.

Most US citizens oppose the idea of living next to data centres

Despite such commitments, public scepticism around AI infrastructure remains high.

A recent Gallup survey found that 71 per cent of Americans oppose the construction of AI-focused data centres in their local communities. Nearly half of respondents said they strongly opposed such developments.

Interestingly, respondents appeared more comfortable with other forms of major infrastructure. Opposition to nearby nuclear power plants stood at 53 per cent, making AI data centres less popular than nuclear facilities among those surveyed.

Only around one in five Americans said they supported having an AI data centre built near where they live.

Environmental concerns emerged as the dominant reason behind the opposition. Around seven in 10 respondents said they were worried about the impact of data centres on local ecosystems and natural resources.

Many cited the enormous electricity and water demands associated with AI facilities, while others raised concerns about pollution, industrial noise and the loss of land that could otherwise be used for housing, recreation or community projects.

The findings underscore a growing challenge for technology companies. As AI becomes increasingly embedded in everyday life, public attention is shifting beyond chatbots and software tools to the physical infrastructure powering them. For firms such as Google, balancing AI growth with environmental responsibility may prove just as important as building the next generation of intelligent systems.