A new study published on the arXiv preprint server highlights the alarming consequences of artificial intelligence (AI) on air pollution. Researchers at UC Riverside and Caltech found that the growing demand for data centers is driving toxic air pollution, with estimated annual costs reaching $20 billion in the US by 2030.
The study reveals that power plants and backup diesel generators supplying electricity to data centers are spewing out pollutants, including fine particles and nitrogen oxides, which can cause premature deaths and severe health issues. According to the researchers, as many as 1,300 people could die prematurely each year from AI-related air pollution by 2030.
The tech industry’s focus on carbon emissions and water usage has left a glaring omission: unhealthful air pollutants. The authors of the study recommend that standards be adopted requiring tech companies to report air pollution caused by their power consumption and backup generators.
Moreover, the study found that low-income communities are disproportionately affected by AI-related air pollution due to their proximity to data centers or power plants. The pollution also drifts across state lines, impacting communities far beyond the data center’s location.
The researchers estimate that if these backup generators emit at their maximum permitted level, the annual cost will increase tenfold, reaching $1.9 billion to $2.6 billion. In some areas, the public health cost associated with AI processing centers exceeds what tech companies pay for electricity.
As AI continues to reshape how we work and play, the study warns that the resulting air pollution is expected to steeply increase. The researchers call for urgent action to address this public health issue, emphasizing that family members of those with asthma or other health conditions may already be affected by the air pollution from data centers.
Source: https://phys.org/news/2024-12-ai-power-demands-toxic-air.html