Data Center Intelligence — Weekly Roundup (Feb. 9–15)
Weekly Data Center Dispatch
Last week felt like the point in the season when the scoreboard stops telling the whole story.
The headline numbers remain enormous: billion-dollar campuses, gigawatts of planned demand, and utility investment programs measured in the tens of billions. But underneath those numbers, a more important question is emerging:
Who carries the cost and execution risk required to make all of this capacity real?
Data center developers need power sooner. Utilities need time and capital to build the supporting infrastructure. Communities want economic development without higher household bills. Technology companies want flexibility without being accused of shifting their costs onto everyone else.
That tension shaped the week.
Below are the stories that mattered, in plain English, with what they mean for operators, customers, investors, and anyone doing FP&A.
1) Industry momentum
Major construction, capital commitments, and the rising price of AI infrastructure
Meta begins construction on a $10 billion Indiana data center
Meta broke ground on a major AI-focused data center campus in Lebanon, Indiana. The project is expected to reach approximately one gigawatt of capacity and begin operating around late 2027 or early 2028.
What this means: Gigawatt-scale campuses are moving from ambitious announcements into active construction. But a project of this size is no longer simply a real estate development. It is a regional infrastructure program involving power generation, transmission, water, roads, labor, equipment, and local government.
Analogy: This is not building a new stadium. It is building the stadium, the surrounding highway exits, the power plant, and an entire training complex at the same time.
Meta is funding the Indiana project directly
Unlike some recent AI infrastructure projects that used joint ventures or third-party financing, Meta indicated that it would fund the Indiana development itself.
What this means: The largest technology companies have several financing paths available. They can use their balance sheets, create infrastructure partnerships, sign capacity agreements, or move projects into specialized financing vehicles depending on the risk and capital strategy.
Analogy: A wealthy franchise can pay cash for the arena, borrow against future ticket sales, or bring in an ownership partner. The building may look the same, but the risk sits in a different place.
The scale of capital deployment is becoming a competitive advantage
Meta has outlined hundreds of billions of dollars of planned U.S. infrastructure investment over the coming years as it competes for AI leadership.
What this means: Infrastructure capacity itself is becoming part of the product strategy. Companies that can secure chips, land, power, construction capacity, and financing may be able to move faster than competitors that have equally strong software but weaker infrastructure execution.
Analogy: Every team may know the same plays. The advantage belongs to the organization that can afford the deeper roster and keep its best players on the field.
The equipment and labor market continues to tighten
Developers, utilities, and independent power producers are competing for many of the same turbines, transformers, engineers, EPC firms, contractors, and skilled trades.
Some participants are reportedly exploring refurbished or repurposed generation equipment because traditional procurement timelines no longer meet project schedules.
What this means: The constraint is not only power generation. It is the physical equipment and trained workforce needed to build and connect that generation.
Analogy: Everyone has the financing to build a new arena, but there are only a few qualified crews available to pour the foundation and wire the building.
Capital is abundant, but execution capacity is not
The week reinforced a basic truth: financing alone cannot manufacture transformers, create experienced electrical engineers, or shorten turbine lead times.
What this means: Investors should distinguish between financial capacity and delivery capacity. A well-funded project can still miss its schedule when the supply chain, grid, labor market, or permitting process cannot support it.
Analogy: Having the largest payroll does not guarantee a championship when there are no healthy players left to sign.
2) Future expansion
Utilities move into the center of the AI buildout
Duke Energy raises its five-year capital plan to $103 billion
Duke Energy increased its five-year capital expenditure plan by approximately 18%, bringing the total to $103 billion. The utility cited growing electricity demand, including substantial new requirements from data centers.
Duke has secured agreements representing roughly 4.5 gigawatts of data center demand, with an additional pipeline of approximately nine gigawatts.
What this means: Data center development is beginning to reshape utility capital programs. The investment required extends well beyond individual substations and can include generation, transmission, distribution, fuel infrastructure, and grid modernization.
Analogy: The home team has become so popular that the city now needs to rebuild the roads, expand the rail system, and upgrade the electrical network around the stadium.
Utility forecasts are becoming data center forecasts
Large-load projects can materially affect utility demand growth, capital expenditures, financing needs, earnings, rate cases, and regulatory strategy.
What this means: The utility and data center financial models are becoming increasingly connected. A developer’s schedule can affect a utility’s investment plan, while a utility delay can move the developer’s revenue date by years.
Analogy: Two teams are running different playbooks, but the success of each depends on the other arriving at the same place at the same time.
Contract structures are being used to protect existing customers
Duke said it is using customized contracts and available tax credits to help manage the infrastructure investment associated with large new loads and reduce pressure on customer bills.
What this means: Utilities and regulators increasingly want data center customers to provide minimum-payment commitments, credit support, longer contract terms, or contributions toward required infrastructure.
Analogy: Before expanding the arena for one major tenant, the owner wants a long-term lease and a guarantee that regular fans will not have to pay for the new luxury section.
The next wave of capacity will depend on coordinated planning
The Indiana and Duke announcements demonstrate that the largest projects require alignment among customers, utilities, developers, government officials, equipment suppliers, and capital providers.
What this means: Development schedules built independently by each participant are likely to fail. Integrated milestone planning is becoming an operating requirement.
Analogy: The quarterback, receivers, offensive line, and coaching staff cannot each use a different clock and expect the play to work.
Power availability is becoming contractual, not conceptual
It is no longer enough to say that a market has attractive power fundamentals. Developers need clear service agreements, infrastructure scopes, cost-allocation terms, construction milestones, and remedies if the schedule changes.
What this means: Power should be treated as a contracted delivery workstream with measurable conditions, not as a general site-selection assumption.
Analogy: A verbal promise that the bus will arrive is not enough when the entire team’s season depends on making the flight.
3) Green energy and environmental builds
The ratepayer question moves to the front of the conversation
Anthropic commits to paying for required grid upgrades
Anthropic said it would offset the costs of grid infrastructure needed to connect its data centers, including through the rates it pays for electricity.
What this means: Large technology companies increasingly recognize that community acceptance depends on demonstrating that households and small businesses will not subsidize AI infrastructure.
Analogy: The VIP suite can be built, but the people in the regular seats should not receive the construction bill.
“Paying your own way” may become the new baseline
Voluntary commitments from large technology companies could influence future utility contracts, legislation, regulatory decisions, and community-benefit agreements.
What this means: Operators should prepare for greater transparency around infrastructure costs, energy rates, contract terms, and how project risk is allocated between the customer and other ratepayers.
Analogy: One team agreeing to pay the full luxury tax can quickly change what the league expects from everyone else.
Grid upgrades are not the same as clean energy
A company can fund the transmission and substation infrastructure required for its facility while still relying on a power supply with a significant carbon footprint.
What this means: Cost responsibility and environmental performance should be modeled separately. Paying for the grid connection solves one problem; determining the source and emissions profile of the electricity solves another.
Analogy: Paying for the team bus does not tell you what kind of fuel the bus uses.
On-site and behind-the-meter generation remain attractive
Long utility timelines and equipment shortages continue to push developers toward natural gas generation, fuel cells, batteries, renewable energy, microgrids, and hybrid systems.
What this means: Behind-the-meter power can improve schedule control, but it introduces fuel, emissions, maintenance, permitting, availability, and capital risks that may not exist in a traditional utility arrangement.
Analogy: Owning the team plane gives you control over the schedule, but now you also own the maintenance, staffing, fuel, and safety obligations.
Sustainability claims are moving toward financial accountability
The strongest environmental commitments increasingly include specific responsibilities: who funds the upgrade, who absorbs the operating risk, what resource is being used, and how performance will be measured.
What this means: Vague sustainability language is losing value. Investors, customers, regulators, and communities want measurable obligations tied to real projects and real costs.
Analogy: The market wants the official box score, not the postgame speech.
4) Government policy that affects data centers
Moratoriums, tax incentives, and the search for a new development framework
New York lawmakers consider a three-year moratorium
Legislation introduced in New York proposed pausing certain new data center developments while the state evaluates their energy, water, climate, and community impacts.
What this means: Even proposals that do not become law can create development uncertainty, lengthen approval cycles, and change how investors assess a market.
Analogy: The league may not cancel the season, but the possibility of a lockout still changes contract negotiations.
Georgia lawmakers move to reconsider data center tax benefits
The Georgia Senate advanced efforts to reduce or eliminate certain data center tax incentives as lawmakers reconsidered whether the economic benefits justify the cost and infrastructure impact.
What this means: Tax incentives that were once treated as permanent parts of a market’s value proposition can no longer be assumed to remain unchanged throughout a project’s life.
Analogy: A team signed a player expecting the salary-cap rules to stay the same. The league is now rewriting them before the contract expires.
Rhode Island considers a stronger regulatory structure
Policy discussions in Rhode Island focused on regulating data center development and evaluating the associated energy and environmental effects.
What this means: States without large established data center markets are trying to write rules before the development wave arrives.
Analogy: Smaller cities are studying what happened in the major leagues before approving their first expansion franchise.
Michigan communities continue adopting temporary moratoriums
Additional Michigan municipalities, including South Lyon and York Township, adopted temporary restrictions while officials studied zoning, power, water, noise, and community-impact issues.
What this means: Local development risk is becoming highly fragmented. A supportive state policy does not guarantee approval from the municipality controlling zoning and land use.
Analogy: The league may approve the team, but the local stadium authority still controls whether it can play.
El Paso moves toward a formal data center policy
El Paso officials prepared to consider a citywide framework for future data center development, including infrastructure, utility, zoning, and community considerations.
What this means: Communities increasingly want a repeatable policy rather than negotiating each project from scratch. That can create more requirements, but it may also produce greater predictability for responsible developers.
Analogy: A written rulebook can be stricter than an informal handshake, but at least every team knows how the game will be called.
What FP&A should take from this week
If you only remember three things for forecasting:
Separate facility capex from enabling-infrastructure capex. The building is only one part of the investment. Generation, transmission, substations, water systems, roads, and community commitments can materially change the total project cost.
Model who pays, not just what it costs. A grid upgrade may be paid by the utility, the developer, the technology customer, or ratepayers. The economic result changes depending on where that obligation sits.
Treat policy assumptions as variables. Tax exemptions, utility rates, zoning approvals, moratoriums, and infrastructure requirements can change between site selection and ready-for-service.
Closing thought
Last week made the industry’s next challenge clearer.
Data centers are no longer operating at the edge of the power system, public policy, or community planning. At gigawatt scale, they are becoming major participants in all three.
That brings opportunity, but it also changes the standard of execution.
The industry must prove not only that it can build enormous facilities, but that it can pay for the infrastructure they require, explain the benefits to the communities hosting them, and manage the risks without shifting them to people who never approved the project.
In sports terms, the industry has moved beyond drawing up bigger plays.
Now every team must show who blocks, who pays for the stadium, who manages the clock, and what happens when the original game plan breaks down.
The winners will be the organizations that combine ambition with cost discipline, infrastructure coordination, and public trust.
Because the next generation of data centers will not be judged solely by how much compute they deliver.
They will also be judged by how responsibly they were built.
The content is based on public information and personal analysis. This is not financial or investment advice.
