LPSC approval advances Entergy's Meta-linked buildout, but cost allocation remains the real fight
This approval is a milestone, not a finish line. LPSC approval of the landmark agreement moves Entergy closer to building new gas plants for Meta's $10 billion AI data center, with two Richland units due late 2028 and Waterford by the end of 2029. The next battle is over who funds the grid infrastructure as power becomes the scarce asset.
What cleared, and what is still contested
Entergy's pitch is straightforward: large data-center customers should carry a heavy share of grid upgrades so existing customers are not left paying for the full buildout. The company says that arrangement helps 2.3 million Entergy customers see lower bills over time. Skeptics are not fully convinced. They argue much information is hidden by confidentiality agreements, that Entergy's renewable commitment is nonbinding, and that residential ratepayers could still absorb risk if project scope or costs change.
That tension matters because power is becoming the first filter in site selection. In an AI-driven buildout, reliable electricity is becoming one of the most valuable infrastructure assets. The key question is whether the utility can keep ratepayer exposure limited while large tenants pay their share.
The core issue is cost allocation, not whether AI needs electricity
How Entergy says the economics work
From a regulatory standpoint, this is a cost-allocation problem. Data centers create large, lumpy load. If a utility builds for that demand without meaningful contribution from the new customer, the fixed costs can spread across the broader customer base. If the large customer pays more of the upgrades, the pressure on residential and small-business bills should be lower.
That is the core of Entergy's case. The company says large data-center customers pay a big share of the grid upgrades and maintenance costs that smaller users would otherwise help fund. It also says the current wave of data-center demand would deliver $7 billion in savings to 2.3 million Entergy customers over the next two decades. If that accounting holds, the project is not just a new plant story; it is a claim that large load can improve the system's cost base rather than simply expand it.
Why the Meta project is the real test case
The Meta campus is the clearest test of that model. Once completed, it will be more than 4-million square feet, and Entergy is preparing three new combined-cycle facilities on a late-2028 to late-2029 timeline. That is large enough to reshape local generation and transmission planning.
If the customer contribution works as described, the fixed costs of new plants, transmission, and reliability upgrades are pulled onto the data center's cost stack. That would leave residential and small-business users paying for less of a dedicated AI power corridor. If the contribution proves thinner than advertised, the burden shifts back toward the broader rate base.
The bear case: transparency and scope risk
The main risk is simple: if the customer contribution is weaker than promised, ratepayers keep more of the risk. Critics argue much information is hidden by confidentiality agreements, that Entergy's renewable commitment is nonbinding, and that key agreements affecting ratepayers are not fully visible. There is also expansion risk: if demand ramps faster than planned or the facility grows beyond its current scope, the original cost-sharing model could come under pressure.
For regulators and investors, the real catalyst is not the press release. It is the regulatory accounting behind cost recovery and customer contributions.
Why Louisiana could matter for utilities beyond this project
Louisiana may become a template for how regulators value large data-center load. What happens here can shape how the market thinks about utilities across Entergy's service areas and across the sector, because AI demand is turning power into a site-selection filter before fiber, land, or incentives power is now becoming the first filter in site selection. If Entergy can show that hyperscale load can help finance the grid infrastructure it needs, investors may start to view utilities less as slow-growth bond proxies and more as essential infrastructure platforms for AI-era power demand.
The pricing logic depends on proof, not narrative
The bullish version of the story is straightforward: when a large, predictable load helps underwrite new capacity, the cost burden does not have to spread evenly across all households. That helps explain why officials have argued data-center demand could ultimately put downward pressure on electricity prices by bringing on new generating capacity and improving the use of backup resources. Entergy says those dynamics are part of the broader case for AI-linked buildouts ultimately drive down electricity prices through the addition of new generating capacity and better use of existing backup power resources.
If that holds at scale, the first utilities to secure anchored hyperscaler demand could be valued more heavily for controlling the infrastructure layer of compute.
What would change the story later
The hinge is scale and transparency. The current Louisiana project centers on a 2GW facility, but reports say Meta is discussing expansion toward 5GW. That would change the economics of every plant, transmission line, and reliability upgrade.
Watch these signals going forward:
- whether customer contributions and cost-sharing terms remain clear as scope evolves
- whether regulators keep the accounting transparent
- whether projected bill impacts still look credible if the facility grows
If those signals turn positive, the market may keep moving toward a framework that sees utilities as AI infrastructure partners. If they turn negative, the story is more likely to stay focused on ratepayer risk.