Cloud Capacity Risk: MA Data Centers Over 25 MW Now Need Local Approval

Every SaaS business depends on someone else's data centers, and the buildout just hit a wall in Massachusetts. With state permits blocked for facilities over 25 MW until developers win local approval, fund clean energy, and sign community benefits agreements, cloud providers and colocation operators face higher siting costs and longer timelines in one of the Northeast's key markets. For SaaS leaders, that means capacity availability, latency, and pass-through pricing risk for East Coast customers — and a signal that infrastructure may shift to friendlier regions.

On September 10, 2026, Massachusetts Governor Maura Healey signed an executive order that fundamentally reorders how data centers can be built in the Commonwealth. The order directs state agencies to withhold "any permits or authorizations" from data center projects with peak electricity demand greater than 25 megawatts unless the applicant first secures local approval, funds its own clean energy, meets water, air, and community engagement standards, and enters into a community benefits agreement. It also imposes a new fee that flows into a ratepayer protection fund designed to return money to Massachusetts ratepayers. The immediate effect is to give municipalities a hard gate before state permitting can begin, converting every qualifying data center proposal into a local political negotiation backed by a formal state stick.

“

On September 10, 2026, Massachusetts Governor Maura Healey signed an executive order that fundamentally reorders how data centers can be built in the Commonwealth.

The 25 MW threshold is the pivotal number. ISO-NE, the region's grid operator, pegs one megawatt as enough to serve 750 to 1,000 average New England homes. A single 25 MW facility therefore represents the demand of roughly 18,750 to 25,000 homes. Healey's stated rationale is grid equity and reliability: "We can't have a situation where data centers are consuming more energy that's available on the grid at the expense of the rest of our businesses and our residents." That framing matters because it explicitly binds data center growth to the health of the shared grid, a theme building across the country as AI and cloud expansion strain regional grids.

The ratepayer protection fund is among the order's most distinctive features. Rather than simply denying projects or capping load, Massachusetts extracts a fee from data center developers and routes it back to retail customers. This is an emerging policy instrument — sometimes called a load-placement fee or customer-protection charge — intended to prevent incumbent ratepayers from underwriting the grid upgrades that hyperscale demand requires. For developers, it converts a social externality into a direct cost that must be modeled into project economics alongside the clean energy self-funding requirement.

The Westfield case gives the policy a concrete local face. Servistar Realties had proposed a 10-building campus built over a decade — the largest data center in Westfield and, by the coverage's account, in New England. Municipal officials initially supported the project, attracted by millions in property tax revenue. But community sentiment turned; city councilors voted earlier in 2026 to enact a one-year moratorium on additional data centers. Although the moratorium did not immediately halt Servistar's project, it signaled a decisive cooling. Under Healey's new order, a project on that scale would now need explicit local approval as a prerequisite for state permits, a sequencing likely to be tested by that very campus or the next one like it.

For investors, developers, and technology operators, Massachusetts has become structurally more expensive and slower for large data center siting. Community benefits agreements — familiar from large real estate and infrastructure deals — are now regulatory conditions rather than voluntary gestures. Clean energy self-funding adds capital expenditure, and the ratepayer fee adds an operating charge. The net effect compresses project margins and extends timelines at precisely the moment AI-driven compute demand is surging. Some hyperscalers may respond by shifting capacity to neighboring states with more permissive regimes, by accelerating on-site generation such as solar-plus-storage, fuel cells, or gas turbines, or by absorbing the costs and passing them through to cloud customers.