Staff Wrote the Rules. Council Wants to Skip Ahead.

Friends & Neighbors,

Fair warning: this one’s a policy deep dive, probably a 10-12 minute read. There’s a real fight brewing at City Hall and I wanted to give it the full treatment instead of a quick take.

On May 15, Mayor Watson posted a proposal on the Council Message Board asking City staff to build a framework for regulating data centers in Austin, joined by Council Members Ellis, Alter, Velásquez, and Siegel. Today, three months later, staff delivered an 11-page memo with their answer. Within hours, the Mayor’s office and Council Member Ryan Alter had already posted follow-ups on the message board pushing for something considerably tougher than what staff recommended. I want to walk you through all three documents, because the gap between them is the actual story here, and because I think there’s a state-level lesson sitting right next to this fight that Austin would be smart to pay attention to.

A data center is a building full of servers. Most of the ones already in Austin are small, unremarkable, and have coexisted with office parks for years without anyone noticing. What’s changed is the emergence of “hyperscale” facilities driven by AI and cloud computing, campuses that can run 100 to over 1,000 megawatts, cover hundreds of acres, and use enough electricity to power tens of thousands of homes off a single site. For reference, the staff’s memo notes that one megawatt serves roughly 250 residential homes. A single hyperscale campus can be the electrical equivalent of adding a small city.

Austin currently has ten data centers on Austin Energy’s grid, ranging from 1 to 20 megawatts. None of them are close to the 75-megawatt threshold Texas law uses to define a “large-load” customer, and Austin Energy says it isn’t aware of any large-load data center currently seeking to connect here. Austin Water reports data centers account for less than 0.3% of the city’s total water use, and none of its data center customers meet the utility’s own “large volume” definition. In other words, Austin doesn’t have a hyperscale data center problem today. It has a real, well-founded fear of getting one before it has rules ready.

The memo, signed by Austin Planning’s Andrea Bates, Austin Energy’s Stuart Reilly, and Austin Water’s Shay Ralls Roalson, does not recommend a ban. It recommends building an actual regulatory structure where essentially none exists today. Right now, Austin’s code doesn’t even have a specific land-use category for data centers; they're lumped under “Communications Services,” the same category as a radio station. Staff wants to fix that with three moves:

1. Create new, size-based land-use categories specifically for higher-impact and hyperscale data centers, and limit them to zoning districts that can actually accommodate the scale, likely closer to heavy industrial than office parks.

2. Require performance standards addressing water use, noise, heat, and lighting, including things like mandatory post-construction acoustical studies, generator testing limited to daytime hours, and enhanced buffers next to homes, schools, and hospitals.

3. Make explicit, as a matter of economic development policy, that Austin will offer zero incentives to attract data centers and won’t exempt them from resource-efficiency standards.

Staff also looked at how other cities are handling this. Fort Worth is developing 250-foot setbacks and acoustic barriers. Irving requires a 300-foot residential setback and caps building height at 75 feet. Frisco only allows them via special permit in highway districts. And San Marcos, down the road from us, went further than anyone: in June, its Council voted 4-3 to ban data centers outright, in every zoning district, citywide.

That last example matters because it’s exactly the fork in the road Austin is standing at right now.

Within hours of the memo posting, Mayor Watson’s chief of staff posted a follow-up on the message board recommending “a freeze on approvals of any application to any City department that is for the purpose of placing a data center” anywhere in Austin, its extraterritorial jurisdiction, or the service area of an Austin utility, until Council adopts a complete policy. Council Member Ryan Alter went further still, posting his own list of clarifications he wants added: prohibiting large data centers in every existing zoning district, refusing to map any new data center zones at all so that every single one would require individual Council approval, and layering in a carbon-free generation requirement on top of everything staff proposed.

I want to be fair to what’s motivating this. Nobody wants Austin to end up like the unincorporated parts of Hays County right now, where residents packed San Marcos City Hall to protest a proposed AI data center before that city’s ban passed. The instinct to move fast and move hard is understandable.

But I think it’s worth slowing down on this point, because Texas itself is running a live experiment on what happens when local governments reach for the strongest tool available first.

On August 3, Governor Abbott directed the Public Utility Commission and ERCOT to conduct a full audit of every data center project in the interconnection queue before any additional ones are approved to connect to the grid. That’s a serious move. ERCOT is reportedly sitting on requests for 474 gigawatts of new capacity, roughly five times the state’s record peak demand, and about 90% of that is data centers. Abbott’s letter states plainly that any project failing the audit “must be denied connection to the Texas grid.”

I want to flag something here that doesn’t show up much in the coverage, but it’s the piece that actually worries me most, given my own background in emergency management: ERCOT nearly failed catastrophically in February 2021. Hundreds of Texans died in a grid failure that most people didn’t think was possible until it happened. The queue Abbott just ordered, which audited more than five times the state’s record peak demand, consists nearly entirely of data centers. That’s not just a market question or an environmental one. It’s a resilience question: does adding this much concentrated, continuous load to a grid that already failed once make the next extreme weather event more dangerous, not less? Austin Energy’s own large-load agreements already require customers over 20 megawatts to curtail at least half their usage during an ERCOT emergency, a real, existing safeguard worth knowing about. But I’d want to know whether the state’s audit is actually stress-testing for a repeat of February 2021, not just checking whether today’s demand forecasts add up on a normal day. That’s the difference between planning for growth and planning for the day the grid gets tested for real.

Here’s the part I think matters most for Austin’s decision: Abbott did not propose a moratorium. His actual policy priorities, which he’s said he’ll pursue with the Legislature when it reconvenes in January 2027, are narrower and more surgical: require data centers to pay for their own electric infrastructure instead of spreading that cost across residential ratepayers, mandate water-efficient cooling, require new data centers to add their own generation capacity rather than just drawing down the shared grid, and set siting and setback standards. That’s a regulation aimed at making sure data centers pay their own way and meet real standards, not a blanket “not here” answer.

Compare that to how blanket local bans have actually fared this year. Hood County commissioners considered a moratorium in February and backed off after state Sen. Paul Bettencourt warned them that counties don’t have the legal authority to do it. Hill County passed the state’s first county-level data center moratorium in May, then rescinded it within weeks after a developer sued for $100 million. San Marcos’s ban, unlike the county attempts, is standing so far, because as a home-rule city, it has real zoning authority that counties don’t, but Bettencourt has already said publicly he intends to challenge it, and the Data Center Coalition has called that kind of local moratorium a signal that a city is “closed for business.”

Austin is a home-rule city, too, which means legally, Austin is positioned more like San Marcos than like Hood or Hill County if it wanted to go that route. But being legally able to attempt the broadest possible ban isn’t the same question as whether it’s the wisest first move, especially when the state’s own regulators are mid-audit, the Legislature’s interim committees are still gathering information with reports due by the end of this year, and Austin’s own staff, after three months of work, concluded the smarter path is defined land uses, real performance standards, and zero incentives, not a freeze on every application in the pipeline before any policy exists at all.

I don’t think Council Member Alter’s substantive asks are unreasonable on their merits. Real setbacks, real noise standards, keeping data centers out of neighborhoods, those are exactly what staff already recommended. Where I get uneasy is the mechanism: refusing to map any zones at all so every project needs individual Council approval effectively turns land use policy into a case-by-case political decision for every future applicant, and a blanket freeze on all applications, city, ETJ, and utility service area, before any actual policy exists, is a much bigger hammer than the memo called for.

There’s a question sitting underneath all of this that doesn’t get asked enough: what does Austin actually get in exchange? The memo notes plainly that data centers “are not large employers.” These are enormous capital investments, hundreds of millions or billions of dollars in physical infrastructure, that add real property tax value but comparatively few permanent jobs relative to their footprint, their water draw, and their electricity demand. I’ve spent this whole month skeptical of the City handing out incentive deals without asking what Austin gets back for it. This memo already takes incentives off the table here, which I think is the right call. But the same question is worth asking even without an incentive attached: a facility that uses a small city’s worth of power and water while employing a few dozen people on-site is a different kind of neighbor than a business that does the reverse, and Austin’s zoning and siting rules should probably account for that difference directly, not just by size, but by what the City actually gets back for hosting them.

Austin doesn’t have a hyperscale data center knocking on the door today. That gives Council the rare luxury of getting this right rather than getting it fast. I’d rather see Austin adopt the specific, defined, standards-based framework already on the table than reach for the most aggressive tool available before it’s actually needed.

Here’s the piece I haven’t seen anyone at City Hall raise, and it’s the one I’d actually push for if I still had a vote.

Even if Austin adopts the toughest local freeze or outright ban Council can legally write, state law can simply override it once the Legislature acts. That’s not hypothetical. Governor Abbott has already said he intends to pursue a formal data center framework with the Legislature when it reconvenes in January 2027, covering exactly the same ground Austin is trying to regulate right now: infrastructure cost allocation, water efficiency, siting standards. Texas has a long, well-established pattern of the Legislature preempting local ordinances outright once it decides a matter is better handled at the state level. Austin has been on the losing end of that fight before, on everything from short-term rentals to paid sick leave to, as I’ve written about this month, its own police budget. There’s no reason to assume data centers get treated any differently once state lawmakers write their own rules.

Given that, I think the smarter first move isn’t a freeze at all. Austin already runs a real, standing process for exactly this kind of moment: once a year, Council adopts a State Legislative Agenda through a resolution that directs the City’s own Government Relations Office on what to advocate for at the Capitol, and whatever ends up as the City’s official position is only whatever actually passes in that resolution. This can be done by the council at any time of the year, through a resolution. “Preserving local authority over zoning, ordinances, and local control” is already one of that program’s stated priority categories. Any Council Member, not just the Mayor, could propose adding data center policy to Austin’s next legislative agenda right now. And if it gets left out of the draft agenda staff brings forward, any Council Member can also offer it as an amendment when that item comes up for adoption. Either path directs the City’s intergovernmental relations team to actually show up at the Capitol and represent Austin’s position: protect the City’s zoning authority over where these facilities can go, make sure the companies themselves cover their own infrastructure costs instead of spreading the bill across residential ratepayers, and set real water and energy standards, before the Legislature writes a bill Austin never had a seat at the table for.

I also don’t think the freeze is actually the faster, simpler option it might look like from the outside. A message board post recommending a freeze isn’t binding on anyone. For a real freeze to take effect, it would need the same basic mechanics as any other Council action: a sponsor, enough co-sponsors to get it on the agenda, and a floor vote, the same lift as the legislative agenda amendment I’m describing above. If that freeze is intended to serve as an actual moratorium, directing staff to stop accepting or processing specific applications, it likely requires more than a simple resolution.

Actions that directly restrict land use and development activity typically require an ordinance, with multiple readings, public notice, and a public hearing, not a same-day vote. That’s exactly the kind of process that invites legal exposure, as happened to Hill County when a developer sued over its moratorium for $100 million. A legislative agenda amendment carries none of that risk. It’s Austin directing its own lobbyists, not restricting anyone’s property rights. If the choice is between two options that both require a floor vote, I’d rather Council pick the one that’s legally safer and gets Austin a seat at the table at the Capitol, not just a defensive posture back home.

That’s the version of this I’d be pushing if I were still on the dais: Adopt the local zoning and performance standards staff already recommended, yes. But also get out ahead of the state process instead of just waiting to see what survives it. A local freeze that might get preempted in five months is, in the end, a weaker position than a City that showed up at the Capitol with a clear, specific ask already on the table.

Council holds a work session on this August 25 and is scheduled to discuss and possibly act on August 27. I’ll be watching to see whether they build on the staff’s actual recommendation, reach for the freeze Mayor Watson’s office is asking for, or do the thing I think would actually work best: both the local rulebook and a real seat at the legislative table before the state writes the rules without them.

Until Next Time,

Mackenzie