Air pollution from data centers used to power artificial intelligence could lead to 1,300 premature deaths annually and public health costs reaching $20 billion by 2030, researchers warn.
Eve Upton-Clark reports for Fast Company.
In short:
- A study predicts AI-driven data centers will produce pollution on par with California's entire vehicle fleet by 2030.
- The energy demands of large language models create fine particles and nitrogen oxides linked to asthma, cancer and other diseases.
- The Department of Energy projects data center energy use could double or triple by 2028, equating emissions from training AI models to tens of thousands of cross-country car trips.
Why this matters:
AI’s environmental impact extends beyond carbon emissions to include significant air pollution affecting nearby communities. Transparent reporting and stronger regulations could help mitigate health risks as AI adoption grows.
