Miami is becoming home to companies building some of the infrastructure behind America's AI boom.
One of them is Hut 8, the Miami-headquartered energy and digital infrastructure company that is developing a massive AI data-center campus in Texas.
Now the company is taking a position on one of the biggest political questions surrounding that buildout: who should pay for the enormous amount of power and infrastructure required to support AI data centers?
Hut 8 says the answer should be the developers.
The company has backed Texas Gov. Greg Abbott's push for data-center regulations requiring developers to cover the power and infrastructure costs associated with their projects while protecting local communities.
The timing is notable.
Hut 8 is developing Beacon Point, a Texas data-center campus with 1,000 megawatts of utility capacity. The project is designed around large-scale AI and high-performance computing demand and has already been commercialized through two 15-year leases totaling 704 MW of contracted IT capacity.
Why Texas Is Rethinking Data Centers
Texas has spent years positioning itself as one of America's most attractive destinations for large-scale data centers.
Cheap and abundant energy, available land and a business-friendly regulatory environment helped drive that expansion.
But the scale of proposed development is now creating political pressure.
Abbott recently ordered a pause on approvals for new data-center projects requiring grid connections while state agencies examine their impact on the state's electricity system. The governor's directive requires greater information about projects' energy and water use, ownership and local impact mitigation.
The concern is straightforward.
AI data centers can require enormous amounts of electricity. Building the transmission, generation and other infrastructure necessary to serve them can also create costs that extend beyond the boundaries of the facility itself.
That has led to a growing debate over whether ordinary electricity customers should ultimately bear any of those costs.
Texas is now trying to establish clearer rules.
Hut 8 Has a Particularly Large Stake
Hut 8 is not commenting from the sidelines.
Its Beacon Point development is one of the company's flagship AI infrastructure projects.
The campus has 1,000 MW of utility capacity, while two 15-year leases cover 704 MW of contracted IT capacity. Hut 8 has said Phase 1 is targeted for delivery in the third quarter of 2027, with Phase 2 targeted for the second quarter of 2028.
The scale is difficult to visualize.
A project measured in hundreds of megawatts is no longer simply another commercial building with a large electricity bill. It becomes part of the region's energy infrastructure planning.
That makes the question of who pays for that infrastructure central to the economics of the project.
The Water Question Matters Too
Power isn't the only issue.
Large data centers also face scrutiny over water consumption, particularly in regions where communities are already concerned about long-term water availability.
Hut 8 has emphasized a different approach at Beacon Point.
The company says the campus will use closed-loop cooling and water sourced from outside the county, reducing the project's reliance on local water resources.
Those design choices matter because the political debate around AI infrastructure is increasingly moving beyond the simple question of whether data centers create jobs and economic activity.
Communities are asking what they consume.
How much electricity?
How much water?
How much new infrastructure?
And who pays for it?
Miami Is Part of the Story
Hut 8 is headquartered at 1101 Brickell Avenue in Miami, making the company a direct link between Miami's growing technology ecosystem and the infrastructure being built to support AI elsewhere in the country.
That gives Miami an interesting position in the national AI infrastructure race.
The city is not simply trying to attract AI startups. Companies based in Miami are increasingly participating in the capital, energy and infrastructure buildout required to run AI at scale.
Hut 8 is one example.
Its business has evolved from its roots in digital-asset infrastructure toward a broader energy and compute strategy, with AI data centers now a major part of its growth plans. Hut 8 has described its development pipeline as extending to thousands of megawatts of potential capacity.
The New Rule for AI Infrastructure
The significance of Hut 8's position may extend beyond Texas.
The economics of AI are increasingly tied to physical infrastructure.
Models need compute.
Compute needs data centers.
Data centers need electricity, cooling, land and transmission infrastructure.
As those facilities become larger, communities and regulators are increasingly likely to demand that developers account for the external costs.
Hut 8's position effectively accepts that premise.
For a company building a 1,000 MW Texas campus, that is more than a theoretical policy argument.
It is a business model question.
If AI infrastructure is going to expand at the speed investors and technology companies expect, developers will have to demonstrate not only that they can secure power, but that their projects can coexist with the communities and infrastructure systems around them.
The AI buildout may be happening at hyperscale. The political question is becoming who pays for it.
