This new article from the NYT is one of the most blatant misrepresentations of the AI water issue I've seen. A ton of pictures and close-ups of faucets run dry and people suffering from drought, with a subtitle saying "When Microsoft opened a data center in central Mexico last year.... Water outages, which once lasted days, stretched for weeks"
Nowhere in the article does it make clear how much of the community's water the data center is using.
Surprise, again, when you dig in, you find out that the data center is using tiny fractions of the region's water, comparable to any other industry in the region.
The data center seems to have a maximum permit of 25 million gallons per year (about 1/4th of a large car factory).
context.news/ai/thirsty-data…
A maximum permit is often way higher than actual use. It's there for worst-case scenarios, because permits are hard to change once you get them. Microsoft claims it only draws water 5% of the year.
news.microsoft.com/source/la…
The area the data center is in draws 25 billion gallons per year if we multiply the population by the government's given water per person number.
es.wikipedia.org/wiki/Zona_m…
ceaqueretaro.gob.mx/en-litro…
So the maximum amount of the region's water the data center is permitted to draw is only 0.1% of the water there.
If this article said "After a factory was built, a region's water demand went up by 1/1000th. Droughts that lasted days now lasted weeks" and then featured a ton of pictures of people suffering from lack of water, I think the average reader would ask "Wait, what? That's clearly not the cause of the water issue then." The authors surely know that they could just look into how the data center compares to any other regional use of water, but as usual they don't, and leave the reader to infer it must be the main culprit. Intentionally misleading.
When I have time I can look into the other region's mentioned, but this is all pretty easy, and you the reader can do the same!
NYT article:
nytimes.com/2025/10/20/techn…