Google, Meta & Microsoft's AI investment ESG impact

The rapid rise of artificial intelligence is creating public health and environmental challenges.

Research from UC Riverside and Caltech estimates that pollution from data centres has led to over US$5.4bn in healthcare costs in the US over the past five years.

These centres, which power AI, consume vast amounts of electricity, much of which comes from fossil fuels, leading to harmful emissions.

The growing health impact of AI energy consumption

AI’s increasing energy can potentially contribute to serious health issues. Data centres require significant power to function, and in many cases, this energy comes from fossil fuels. The emissions from these fuels have long been associated with respiratory diseases, cancer and other severe health conditions.

According to research by UC Riverside and Caltech, air pollution from data centres cost the US healthcare system an estimated US$1.5bn in 2023 alone, a 20% rise from the previous year.

The study identifies Google, Microsoft and Meta as the biggest contributors to these costs, with Google alone responsible for US$2.6bn over the five-year period.

As AI continues to expand, the energy demand from data centres is expected to grow. In 2025, Microsoft, Google, Amazon and Meta are projected to invest a combined US$320bn in AI infrastructure, more than double the US$151bn spent in 2023.

OpenAI and SoftBank have also announced plans to invest $500bn in AI development.

The International Energy Agency warns that data centre electricity consumption could double by 2026. A single ChatGPT query requires almost ten times the electricity of a standard Google search. Goldman Sachs predicts that by 2030, data centres could consume 10% of all electricity in the US, up from 4% in 2023.

This trend is not limited to the US. In Europe, electricity demand is expected to increase by 40-50% over the next decade, largely due to the expansion of data centres. In Ireland, these facilities now use 21% of the nation’s electricity—more than the combined consumption of all urban households.

Water usage and electronic waste are growing concerns

Beyond carbon emissions, AI’s environmental impact extends to water consumption and electronic waste. Data centres generate significant heat and require extensive cooling systems to function. Many rely on water-based cooling, which has driven up water demand as data centre growth accelerates.

Google’s 2024 Environment Report reveals that its data centres have increased water consumption by nearly 88% since 2019. In drought-prone areas like California, this has raised concerns about long-term water sustainability.

Another major issue is electronic waste. AI chips and servers have limited lifespans, and their disposal can create hazardous waste. Many of the materials in data centre hardware contain toxic substances that pose health risks if not handled properly. As AI infrastructure expands, addressing these environmental impacts is becoming increasingly urgent.

Low-income communities bear the brunt of pollution

The health and environmental costs of AI are not evenly distributed. Data centres are often built in regions with cheaper land and fewer regulations, such as West Virginia and Ohio. These areas typically have lower-income populations that suffer the most from the resulting air and water pollution.

“Unlike carbon emissions, the health impacts caused by a data centre in one region cannot be offset by cleaner air elsewhere,” says Shaolei Ren, an Associate Professor at UC Riverside.

Tech companies often seek to offset their carbon emissions through renewable energy credits, but this approach does not directly reduce pollution affecting local communities. The reliance on market-based solutions rather than actual reductions in fossil fuel use has drawn criticism from environmental experts.

“There are so many issues around carbon credits from an ecological perspective,” says Sebastián Lehuedé, an Assistant Professor at King’s College London.

“If you consume water somewhere to the point where it affects biodiversity in one area, that cannot be offset by having a nice project elsewhere. You’re going to cause irreversible damage if you keep to that logic.”

Moving forward sustainably

Some companies are making efforts to reduce their environmental footprint. Microsoft has signed an agreement with Constellation Energy to restart a nuclear power unit at Pennsylvania’s Three Mile Island in an attempt to lower reliance on fossil fuels.

“To achieve our goal of becoming carbon negative by 2030, we will need a broad range of innovative carbon-free energy solutions,” says Melanie Nakagawa, Chief Sustainability Officer of Microsoft. “Advanced nuclear energy and fusion energy are included in our multi-technology approach to reaching this target.”

Researchers are also exploring ways to create more energy-efficient AI models, improve cooling systems and develop computational methods that reduce power consumption. However, experts warn that as long as AI expansion is driven by profit, sustainability may remain a secondary concern.

Public protests against data centre expansion are increasing in the US, Ireland and Mexico. Concerns over water shortages, pollution and rising energy costs are fuelling local opposition, with campaigners calling for stricter oversight of the tech industry.

“We are going to reach a tipping point where the increasing cost of data and hence, AI, is not just environmentally expensive but also socially expensive,” says Revathi Kollegala, a Digital Strategist at CIFOR-ICRAF. “This will undermine the logic that AI can democratise access to knowledge and reduce inequity. We may have reached that point already or will very soon.”

Explore the latest edition of AI Magazine and be part of the conversation at our global conference series, Tech & AI LIVE. 

Discover all our upcoming events and secure your tickets today.

AI Magazine is a BizClik brand