The Factory Behind the AI Factory: Alabama Is Chasing a $434 Million Prize Few People Are Talking About

Birmingham is already attracting enormous AI infrastructure projects. Now Alabama is competing for something potentially more important to the long-term economy: a factory that would build the machines that keep data centers alive.

Everyone wants the data center.

Cities fight over them. Governors announce them. Utilities redesign grids around them. Developers assemble hundreds of acres for them.

But there is another side of the artificial-intelligence boom that gets much less attention.

Someone has to build the equipment inside them.

And that may be where Alabama’s next big opportunity is emerging.

Carrier Global, one of the world’s largest heating and cooling companies, is considering Alabama for a major new manufacturing operation dedicated to equipment for data centers. The project is also being pursued by Fort Worth, Texas, and a location in Virginia.

The numbers attached to the competing Fort Worth proposal are striking: at least $433.8 million of investment, a massive industrial facility and 495 or more jobs by 2029. Those jobs would carry an average annual salary of at least $75,000.

If built at that scale, it would be the largest facility investment in Carrier’s history.

But the most interesting part of this story isn’t $434 million.

It’s what the factory would actually make.

AI Has a Heat Problem

Artificial intelligence may feel weightless.

A question disappears into the cloud. Seconds later, an answer appears.

Behind that illusion sits an extraordinary amount of physical infrastructure.

Buildings.

Electrical substations.

Transformers.

Generators.

Miles of cable.

Thousands upon thousands of computer processors.

And heat.

A lot of heat.

The more computing power engineers pack into modern AI servers, the harder it becomes to keep those machines at safe operating temperatures. Cooling is no longer a relatively mundane component of a data center. It is becoming one of the fundamental engineering constraints on how much AI computing can be placed inside a building.

Carrier sees the opportunity.

Its commercial data-center portfolio now stretches across chillers, air-handling systems, coolant-distribution units, controls and thermal-management technology designed specifically for enormous computing facilities.

And demand is moving extraordinarily fast.

Carrier reported in July that its data-center orders had increased more than 300% during the second quarter of 2026. Commercial HVAC orders overall rose approximately 65%.

That helps explain why a company best known by many Americans for the air conditioner outside their house is contemplating spending hundreds of millions of dollars on another American factory.

It isn’t really an air-conditioning story anymore.

It is an AI infrastructure story.

Now Look at Birmingham

This is where the timing becomes particularly interesting.

There is currently no public evidence that Carrier’s proposed Alabama factory would be located in Birmingham.

There is also no disclosed connection between the project and either Nebius or QTS.

Those distinctions are important.

But it is impossible to look at what is happening around Birmingham without seeing the beginnings of an unusual industrial pattern.

In Oxmoor Valley, Nebius is developing its BHM01 AI factory on the former Regions Bank operations property at 201 Milan Parkway.

Nebius says more than 1,000 skilled construction jobs will be involved in building the project, with construction continuing through 2028. Its plans include sophisticated closed-loop cooling infrastructure, chillers and dry coolers engineered to support an enormous concentration of computing hardware.

A few miles to the west, another project is taking shape around Bessemer.

QTS, the Blackstone-backed data-center giant, emerged this summer as the company behind the enormous Project Marvel site after an affiliate closed on the property.

These aren’t ordinary office buildings.

They are pieces of the physical AI economy.

And that raises a bigger economic-development question for Alabama:

Why stop at the data center?

The Real Prize May Be the Supply Chain

The first phase of America’s data-center boom has largely been a land-and-power competition.

Who has enough acreage?

Who can deliver hundreds of megawatts?

Who has transmission capacity?

Who can get permits approved?

The next competition may be over the industrial supply chain that follows.

Data centers require enormous quantities of specialized electrical and mechanical equipment: transformers, switchgear, generators, cooling systems, pumps, controls, liquid-cooling hardware and other components.

Those pieces have to be manufactured somewhere.

A Carrier plant would move Alabama one step deeper into that value chain.

Instead of simply providing land and electricity to the companies operating AI computers, the state could manufacture some of the equipment required to build AI infrastructure across the country.

That’s a very different economic proposition.

A hyperscale data center can represent billions of dollars of capital investment while employing surprisingly few permanent workers once construction is finished.

A manufacturing plant works differently.

Factories require technicians, machinists, engineers, supervisors, maintenance workers, logistics workers and managers every day.

If Carrier’s Alabama proposal resembled the Fort Worth version, roughly 500 permanent jobs averaging around $75,000 would represent a substantial recurring payroll.

And then there are the suppliers, truck movements, service contractors and secondary businesses surrounding a manufacturing operation.

For apartment demand and the broader local economy, that can matter enormously.

Birmingham Already Knows This Formula

Birmingham did not become an industrial city because it consumed iron.

It became an industrial city because it sat where the ingredients for an industry came together.

Iron ore.

Coal.

Limestone.

Railroads.

The city made something from them.

More than a century later, the ingredients are different.

Electricity.

Fiber.

Land.

Water infrastructure.

Advanced cooling.

Computing hardware.

But there is an interesting echo of Birmingham’s past in the economic opportunity emerging today.

The cities that win the AI economy may not simply be the places where servers are installed.

They may be the places that learn to manufacture, maintain and support the enormous physical machine surrounding those servers.

That is why the Carrier decision deserves considerably more attention than another generic factory-site competition.

Alabama Has Not Won It

There should be no victory lap yet.

Fort Worth is competing aggressively.

Its proposed Carrier site is a roughly 1.2-million-square-foot industrial building in the AllianceTexas development, and local officials are considering a seven-year tax abatement worth approximately $10.9 million. Carrier job advertisements have already appeared using the Fort Worth address even though the incentive process remains incomplete.

Virginia is also in contention.

Carrier is expected to make a site decision during the third quarter of 2026.

And crucially, the Alabama site itself has not publicly surfaced.

It could be Birmingham.

It could be somewhere else in the state.

Until Carrier or Alabama officials disclose the location, connecting the factory directly to Birmingham would be speculation.

But the timing is nevertheless difficult to ignore.

Alabama is competing for a factory designed to serve one of the fastest-growing parts of the data-center industry at precisely the moment Birmingham is becoming one of the Southeast’s more unexpected destinations for gigantic AI infrastructure projects.

That may be coincidence.

Or it may be an early glimpse of what comes next.

From Data Centers to an AI Industrial Cluster

The best outcome for Birmingham would not be one enormous Nebius project.

Nor would it be QTS alone.

The more consequential outcome would be the creation of a cluster.

Data centers arrive first.

Then engineering companies.

Cooling specialists.

Electrical contractors.

Equipment manufacturers.

Maintenance operations.

Fiber companies.

Power-infrastructure suppliers.

Training programs.

Specialized construction firms.

Eventually, workers and companies begin choosing the region because the rest of the industry is already there.

That is how economic clusters become difficult for competing cities to reproduce.

Carrier has not yet chosen Alabama.

And Alabama has not told us where it wants to put Carrier.

But this particular site-selection contest is worth watching closely.

Because if Alabama wins, the story may be bigger than 495 jobs or a $434 million factory.

It would suggest that the state’s emerging AI economy is beginning to move outside the walls of the data center itself.

And that is where things start getting really interesting.