Solmar Insights
The accelerated growth of artificial intelligence infrastructure is reshaping traditional data center development across the United States, as operators and investors confront mounting challenges related to energy access, site selection, and permitting. At the AI Infra Summit 2026 in Santa Clara, sector leaders spotlighted the complex planning now required to support large-scale AI compute and data movement while balancing regulatory and community demands.
Key figures
Summit held September 15 to 17, 2026, Santa Clara, CA
Focus on gigawatt-scale AI data centers
Utility interconnection, permitting, and community impact highlighted
Power supply and permitting hurdles
The expansion of energy-intensive AI data centers is placing unprecedented strain on grid resources and shifting operator strategies around project siting. An essential takeaway from the summit was that conventional site selection, focused historically on land and capital, no longer suffices. Developers now face elongated utility interconnection queues, variable local regulatory requirements, and complex permitting processes that can delay or upend builds before breaking ground.
Industry panels led by Ilissa Miller, CEO of iMiller Public Relations, together with DataBank CEO Raul Martynek, emphasized that project viability increasingly hinges on early engagement with utilities, regulators, and local governments. Land acquisition is intertwined with the ability to secure timely power delivery and navigate local politics, especially as AI cluster loads approach gigawatt scale and community expectations for transparency rise.
These issues are redefining the risk and cost profile of data center development. Some regions remain more capable of offering scalable grid and permitting support, steering developers toward secondary or tertiary markets where coordinated planning with utilities can make the difference between project advancement and costly delay.
Coordination across design and investment
Panelists at the AI Infra Summit stressed that the race to provide AI capacity at scale now requires a far more integrated approach to infrastructure, spanning site readiness, energy procurement, and capital deployment. Instead of a linear build process, owners and operators align infrastructure design and energy planning in tandem with financing and permitting strategies.
Developers are compelled to assess a site not just for available acreage, but for long-term grid access, physical and digital infrastructure readiness, and pathway to regulatory approval. This cross-disciplinary integration enables earlier risk identification, reducing the likelihood of stranded capital and construction slowdowns. As gigawatt-scale projects proliferate, institutional capital flows increasingly reward those entities that can demonstrate coordinated decision-making with utilities and local officials during site evaluations.
These new requirements are raising the bar for project underwriting, often reweighting site selection criteria in favor of states and municipalities with transparent utility planning, predictable permitting, and a track record of data center engagement. Capital providers pay close attention to early evidence of multi-agency collaboration to de-risk timelines.
Technological shifts in AI compute
The drive for ever larger and more dynamic AI workloads, ranging from advanced large language models to real-time agentic inference, demands technology stacks that exceed conventional data center architectures in both performance and efficiency. Ian Buck of NVIDIA outlined the need for infrastructure built purposefully around AI, stressing the right matching of compute horsepower with networking, memory, and storage to address shifting AI requirements.
Summit discussions reflected a fundamental rethink, moving from one-size-fits-all facilities toward specialized architectures leveraging dedicated hardware, distributed compute, and highly networked environments. This transition impacts not only equipment procurement but also power provisioning, cooling, and spatial planning, particularly as AI clusters utilize ever more concentrated rack densities and power draw per square foot rises.
Greater integration of hardware and system design offers route to cost containment without sacrificing capacity. However, these bespoke environments add to project complexity and interdependency between energy procurement, facility design, and supply chain schedules. The ongoing recalibration challenges site, grid, and capital planning alike.
Data movement and network demand
AI infrastructure buildup is not only about compute power but also hinges on the ability to move vast quantities of data efficiently within and between sites. Summit participants highlighted that scalable, low-latency networks and robust data routing architectures are now as critical as power availability to sustaining AI loads at scale.
The requirement for rapid, seamless data movement inflates demand for high-bandwidth networking fabrics, specialized interconnects, and fiber routing to regional aggregation hubs. The interdependence between networking, data storage, and compute cores drives new investments in both on-premises and cloud-integrated digital backbones, raising the bar for resiliency and physical security in tandem with throughput and uptime.
Infrastructural bottlenecks at the networking or interconnection stage can undermine the utility of even the most powerful AI compute clusters, prompting investors and operators to weigh network proximity, rights of way, and redundancy as critical factors in project feasibility assessments.
Political and community impacts
As the network of AI-ready data centers expands, local political environment and community engagement play an ever greater role in development outcomes. Projects must meet the expectations of host communities, many of whom now demand sustainable energy sourcing, transparency on water and energy usage, and economic contributions through jobs or infrastructure offsets.
Political factors, from state-level incentives to municipal zoning board reviews, inject new variables into development timelines and permitting paths. The inability to demonstrate community benefits or comply with local regulatory priorities can slow or even halt projects, introducing new forms of risk for institutional buyers and developers focused on timely delivery of AI capacity.
Ultimately, firms that prioritize early, continual engagement with permitting authorities and community representatives are better positioned to mediate opposition and streamline approvals, enabling faster capital deployment and realization of revenue streams from AI tenants.
What this means for buyers
Data center land, power, and interconnection capacity in the United States are subject to more stringent site planning, utility coordination, and permitting scrutiny. Project viability now depends less on land and capital alone, and more on early-stage power delivery agreements and community engagement identified at the September 2026 AI Infra Summit. Buyers this quarter should prioritize sites with utility and permitting clarity, and evidence of coordinated planning, to reduce delays in AI-ready capacity delivery.
Reporting via the original publisher
