I was trying to figure out how a datacenter (which doesn't really produce any chemicals) would cause cancer. FTA:

Amazon’s hulking data center, thirsty for water to cool its blazing hot computer chips, supercharged this process, adding millions of gallons of wastewater a year **to the heavy volume of farm runoff, ** which Morrow County was already struggling to keep up with. Soon even the deepest reaches of the local aquifer were tainted, according to RS, as huge volumes of data center and agricultural wastewater saturated the water table.

So the problem is actually farm runoff, but the datacenter evaporation concentrates it. Not quite Amazon's fault for the original contamination, however Amazon could and should absolutely switch to a "Closed Loop" cooling system which would mean the water consumption (used for cooling) would dramatically fall, and this concentration issue of farm pollution would disappear.

But why would they ever? The water out of the aquifer they probably just get for free. That's likely one of the reasons why they built there in the first place.

The unfortunate reality is that we often don't know what's actually causing the cancers, and without a concrete theory as to the mechanism of how the disease is being caused, we don't necessarily even know that the data centre is responsible. When there's an unexplained cluster of rare cancers that are correlated to a suspected cause, then that provides a starting point for investigation at least.

The really grim thing is that I have no trust that this will actually be investigated in a reasonable way. Big corporations get to shove their "externalities" onto us, and escape accountability through plausible deniability. Past precedent means that even if I can't understand how a data centre could be causing cancer, it feels more likely than not that something dodgy is going on. I'll hold this view until this issue is thoroughly investigated in a transparent manner

The unfortunate reality is that we often don’t know what’s actually causing the cancers, and without a concrete theory as to the mechanism of how the disease is being caused, we don’t necessarily even know that the data centre is responsible. When there’s an unexplained cluster of rare cancers that are correlated to a suspected cause, then that provides a starting point for investigation at least.

Thats true. Additionally Amazon (and other companies) have hundreds of data centers scattered around the world. If a regular practice of building and running a datacenter would be causing cancer clusters we'd see that in nearly every case, and we're not seeing that.

The really grim thing is that I have no trust that this will actually be investigated in a reasonable way.

Well, eventually it might be investigated after decades of damage and deaths, and the penalty will be a rounding error for a day's profits from the true offending multinational company. Or better yet, the offending company will cease to do business in the decades of investigation and there will be no justice served for the victims. Hollywood will produce a riveting drama featuring that day's leading man or woman actor and our heartstrings will be pulled long enough until the credits roll, and we move onto the next form of entertainment.

But why would they ever? The water out of the aquifer they probably just get for free. That’s likely one of the reasons why they built there in the first place.

Its not free, but it is much cheaper than running a closed loop water system. There are two datacenters being built in my water co-op's service footprint right now. One is an Amazon AWS datacenter and the other is national brand of server colocation services. The AWS datacenter is using an "open loop" cooling method, just like the DC in the article. It will be a massive water consumer. The other is a "closed loop" cooler, which barely ranks differently than other businesses in the area as far as water consumption.

Amazon is paying our co-op handsomely for the water our co-op is providing, but its still less money than the equivalent costs than a closed loop system and the energy it would take to achieve the same cooling result.