By editorial board • 

What lies beyond the public disapproval of data centers?

Oregonians — ninth among states with about 900 megawatts of commercial colocation data-center capacity — are taking action to slow the growth of massive data campuses. However, as demand for advanced intelligence capacity expands exponentially, those data-center moratoriums may need to be replaced with better strategies to limit the negatives while allowing the inevitable expansion of computer data centers.

In late July, Hillsboro enacted a 120-day pause on applications for new or expanded data centers, perhaps stretching to six months. Last week, Woodburn issued a state-required notice for launching a moratorium process, with Salem following suit this week.

In Yamhill County, despite there being no known plans for such a development, members of the public are already speaking out against the idea of local data centers. McMinnville resident Maggie Bowman has addressed both city councilors and county commissioners requesting bans. And a Wednesday News-Register article on the topic quickly garnered several online comments from detractors. As one commenter argued, “Once a data center is allowed into our beautiful farming and vineyard community… It will be GAME OVER.”

Opponents of escalating data-center development point to costs of new power generation infrastructure that otherwise wouldn’t be needed. They warn against rising power costs, tax subsidies to developers, and competition to power needs for homes, transportation and industry — and there are myriad reports that substantiate those fears.

Data center proponents talk about financial transformation of rural revenue bases from properly taxed centers, permanent well-paid jobs, customers for new alternative power generation projects and service to the growing demand for AI capacity.

The Avison Young U.S. data center report for 2026 Q2 reports a nearly 80% growth in “U.S. inventory” in the past year, i.e. the total electrical capacity of commercial data centers renting computer space and power to multiple customers. Oregon’s 900-megawatt capacity naturally lags far behind the nation’s leaders, including Texas at 14,700 megawatts, Virginia at 7,600 MW, and Georgia, Arizona and Illinois ranging from 3,700 to 2,400 MW.

Consider this analogy: Once automobiles became essential to American life, gas stations — with all their unsightly design and environmental risks — were not to be denied. Resulting regulations were required for admittedly smaller-scale developments, and that is a sequence being replicated with AI today.

Oregonians, and others worldwide, will welcome technology that facilitates smaller-scale data centers that incorporate power generation facilities to serve the surrounding communities. Those centers will need to pay the costs and risks for electricity, water and required public infrastructure.

For now, there is a collision between rapidly growing AI demand and public resistance to massive, hyperscale data hubs. Perhaps this irresistible force crashing into an immovable object will produce hybrid systems of huge regional centers for the highest-capacity computing, augmented by smaller local centers that incorporate community microgrids, distributed power generation and usable heat recovery.

Americans are not likely to give up the underlying infrastructure needed to facilitate exploding demand for AI. We asked one AI platform to speculate for us about possibilities for future integrated community models of data centers, producing this response:

“A future community installation might combine a 10–100 megawatt data center, local generation, batteries and thermal storage, grid connection in both directions, interruptible computing workloads, waste-heat recovery, reclaimed-water or largely waterless cooling, and backup capacity available during community emergencies.

“During normal operation, most power would support the data center. During extreme weather or a grid emergency, some computing could be postponed or shifted elsewhere, allowing generation and storage to support essential community loads.

“That is technically more realistic than imagining electricity being permanently divided between ‘the data center’s power’ and ‘the community’s power.’ The value would come from coordinated operation of a local microgrid.

“The bargain might be: The community provides land, permits and an operating environment; the data center finances enough generation, storage and grid infrastructure to serve itself while materially improving community reliability and affordability. That could be a strong public-benefit model—if community access to the power were contractual and enforceable.”

The central question remains: Not whether data centers produce benefits, but whether Oregon can retain enough of those benefits to compensate for the extraordinary consumption of resources.

Comments

@@pager@@
Web Design and Web Development by Buildable