Spending plans for data centers are surging as the artificial intelligence buildout accelerates, but public resistance to new facilities is emerging as a potential bottleneck for the next wave of energy solutions. A report by consultancy McKinsey & Co. estimates global data center spending could reach $7 trillion by 2030, yet polling cited by Gallup suggests a large majority of Americans oppose local data center construction.
That tension matters for investors tracking small modular reactors, or SMRs—technology often pitched as a faster way to add nuclear power capacity near electricity-hungry developments. At the same time, broader shifts in the power mix, including increased interest in natural gas as a near-term bridge, could further affect the timeline and economics of SMR adoption.
Key takeaways
- Data center buildout outlook: McKinsey estimated global data center spending could reach $7 trillion by 2030, underscoring demand for new power generation.
- Catalyst and risk: Gallup polling found around 70% of Americans oppose building data centers locally, with many citing water, energy use, and environmental concerns.
- Implication for SMRs: If construction slows or delays, the incentive to pay a premium for SMR capacity may weaken.
- Power-market angle: Investors may also watch whether AI load growth is met faster with natural gas while conventional nuclear projects proceed over longer horizons.
What McKinsey sees driving data center spending
McKinsey & Co. linked the data center buildout to what it described as “the race to scale AI,” which is prompting large infrastructure investments. The consultancy’s estimate of global data center spending reaching $7 trillion by 2030 points to a multi-year expansion cycle that is likely to keep electricity demand in focus for utilities, power developers, and grid operators.
However, demand-side growth does not automatically translate into faster construction timelines for new energy infrastructure. Investors are increasingly attentive to whether power supply projects can secure siting approvals and public acceptance—especially when data centers themselves face mounting local opposition.
Public backlash as a potential headwind for SMR timelines
According to Gallup polling cited in the report, about 70% of Americans oppose building data centers in their local areas, including nearly half who are “strongly” opposed. The same polling indicates that opponents frequently cite concerns about resource use—particularly water and energy—along with environmental and related issues such as pollution and noise.
The relevance for SMR investing is straightforward: SMRs are typically marketed as quicker to deploy than traditional nuclear plants, potentially offering capacity tailored to the pace of AI-driven demand. But if data center development is slowed by community resistance, the urgency to pay for near-term nuclear capacity could ease. The article notes that on a per-megawatt basis, SMRs are generally more expensive than conventional nuclear plants, meaning their business case depends heavily on the value of time.
In that framework, a slower data center construction pipeline could push electricity buyers toward alternatives that better match their schedules and costs—even if those alternatives are less “directly nuclear” in the near term.
Natural gas demand could reduce the urgency for faster nuclear buildout
Another factor highlighted is the possibility that AI and related compute operators may meet near-term power needs with natural gas. The article points to work described on Aug. 27, including steps by Space Exploration Technologies to invest more aggressively in natural gas supplies to fuel rockets and support compute infrastructure, even including plans to build its own natural gas pipeline.
The underlying investment argument is that if AI firms can satisfy near-term capacity additions with natural gas, gas could serve as a bridge until larger, conventional nuclear plants come online. The article also cites a plan by Alphabet—parent company of Google—to build three new conventional nuclear power plants in the U.S., each capable of producing at least 600 megawatts.
That dynamic could reshape the competitive landscape for SMRs. If longer-horizon conventional nuclear projects absorb the bulk of incremental demand, SMRs may face a longer wait for the “speed premium” that often underpins their economics.
Where investors may look next
With data center expansion projected to remain a central theme, investors may want to monitor whether local permitting and community sentiment translate into delays—or whether operator engagement mitigates backlash. On the power side, watching how quickly AI-related load is contracted and whether additional capacity additions skew toward gas versus conventional nuclear could help clarify the pace of SMR adoption.
Key near-term signals to follow include policy and permitting developments around data center projects, utility planning for incremental generation, and progress on conventional nuclear projects referenced in corporate and regulatory updates.
