The following is the opinion and analysis of the writer:
Andrew Haines
Last week, the news broke that the Tucson City Council had voted (6-1) to adopt strict and new zoning rules for data centers. These new measures limit the noise and height of any proposed facility. The measures also require developers to demonstrate access to adequate supplies and disclose their energy mix. Many voices in the crowd wanted a total ban on data centers. I cannot blame them for disliking data centers: they can be large, loud, and can use up vast amounts of water and electricity. The council did the right thing to reject a ban and leave the door open for data center construction.
Let’s acknowledge both the urgent need for data centers and the concerns surrounding them, then find a way to build them responsibly.
AI isn’t going anywhere. New and better models are being created every day. The sheer computing power is allowing workers to do what used to take a week in just a few minutes. AI is increasingly embedded in research, medicine, manufacturing, national security, and everyday business. AI also helps those looking to do harm. The recent hack on water infrastructure in twelve states highlights the need for our own cybersecurity computing defense and its backbone: data centers.
People are also reading…
Recent reporting tells us that data center builders are acquiescing to municipal demands in order to get these built. To put another way: the AI industry is so desperate for computing power that they will adhere to Tucson’s rules. Having rules to follow actually provides great clarity. Tucson should not reject a facility that fully powers itself with clean energy, funds necessary grid upgrades, and uses closed-loop cooling to minimize ongoing water loss.
Meta recently negotiated a data center in Louisiana the right way. Meta was reportedly transparent and negotiated in good faith; it met their environmental requirements, and local teachers ended up with bonuses up to $50,000 due to an influx of tax revenue from the data center.
While I wouldn’t expect teacher payouts, this is the model of cooperation that Tucson should follow, because our community needs data centers.
One thing that is rarely mentioned in the data center debate is latency. Latency refers to the time it takes for data to travel from a user's device to a data center and back. The closer the data center is to the user, the lower the latency and the more effective AI computing can be. This matters for real-time applications such as autonomous systems, advanced manufacturing, defense technologies, and medical response tools. Proximity to computing resources can therefore strengthen industries that depend on immediate data processing and give Tucson a competitive edge in attracting high-value, time-sensitive work. Don’t you think Tucson businesses want higher computing power? What about researchers at the University of Arizona?
My prediction: The cities that dominate the coming AI economy will be the cities that have the lowest latency, and the cities with no data centers and high latency will fall into decline. The brightest minds and most innovative companies will go where their talents will be put to best use, where they can efficiently do the best research, where computing power is high, and latency is low.
Tucson does not need to choose between unchecked development and a blanket ban. The council has determined the rules; now let’s build data centers for the benefit of the community and without harm to our beloved environment.
Follow these steps to easily submit a letter to the editor or guest opinion to the Arizona Daily Star.
Haines lives in midtown, believes Tucson is a hidden gem, and frequently uses AI to fact-check his wife.

