Introduction

The Arctic is of critical concern to the North Atlantic Treaty Organization (NATO). The opening of new shipping routes and the increased potential for developing natural resources have made the North a geopolitical hotspot. Adversaries like Russia and China are actively strengthening their defense activities in the region, posing a continued threat to the alliance, including its seven Arctic members. In response, NATO has bolstered its strategic defense actions in the region by incorporating Finland in 2023 and Sweden in 2024 and establishing the Arctic Sentry initiative in February 2026 (NATO 2026). Continued and improved resilience of NATO in the Arctic depends on the critical infrastructure supporting collective deterrence activities.

In this brief, we focus on Alaska, America’s Arctic state, and its infrastructural preparedness to support national and collective security in the North. We focus on infrastructure that can be used for both civilian and military purposes, specifically small modular reactors (SMRs) and data centers. SMRs have been gaining traction and bipartisan support in the United States (Shonerd 2026). At the same time, data centers’ presence in the state has grown alongside the growth of the artificial intelligence (AI) industry.

Some communities on the ground have pushed back on SMRs and data centers, and we argue that the build-up of critical infrastructure in Alaska cannot proceed without robust and consistent consideration of communities’ acceptance. Alaska could contribute to NATO’s resilience by developing systems that are embraced by all stakeholders who coexist within and around the infrastructure, which will require establishing a wide network of trust with those stakeholders. We recommend constructing critical infrastructure only with continuing approval from the necessary stakeholders. Additionally, we recommend that, for continued energy resilience, decision-makers need to ensure that the infrastructure on the ground can sufficiently support modern technologies, including data centers.

Definitions

In this brief, we use the Arctic Council’s Arctic Monitoring and Assessment Programme’s definition of the Arctic region: marine and terrestrial areas above the Arctic Circle at 66°32'N, areas north of 62°N in Asia, areas north of 60°N in North America, marine areas north of the Aleutian Islands, Hudson Bay, and certain areas of the north Atlantic Ocean, like the Labrador Sea (GRID-Arendal 2013). We define a data center as a physical space containing information technology infrastructure meant for processing services and products (Susnjara et al. n.d.). We define AI as a computer system capable of performing tasks with human-like intelligence (Stanford HAI n.d.). We define a large language model (LLM) as a program that is trained on datasets to understand, process, and generate human-like text (Cloudflare n.d.). An SMR is an advanced nuclear technology that is modular and has a capacity of up to 300 MW(e) (Liou 2023). We define dual-use infrastructure as infrastructure that serves both civilian and military purposes (RANE 2025). A microgrid is a self-contained power system that manages its own energy supply and demand and can either feed into the main power grid or run independently as a self-sustaining grid (Ton et al. 2012). Social license to operate (SLO) is the “ongoing approval” of communities and stakeholders involved in a project (IAEA 2019). Energy security, which includes energy resilience, is the possession of uninterruptible, available, and affordable energy with minimal economic, environmental, social, and geopolitical consequences, accounting for disruptions and changing environments (Leddy et al. 2024).

Alaska’s Environment and Infrastructure

Before discussing the need to develop critical infrastructure in Alaska and the Arctic, we need to present the current physical conditions in the state. Increasing resource and infrastructure development, spanning beyond the road system, is shaping Alaska’s vast terrain (Eschenbach et al. 1989).

Alaska’s electric grid is not connected to that of the contiguous United States. The state’s “grid” is a decentralized network of regional, micro, and islanded grids (Peterson et al. 2026). Consequently, Alaskans are vulnerable to volatile fuel prices and the failure of single-source systems (Holdmann 2025). Many remote communities’ power loads range from 100 kW to 5 MW (ACEP 2024), and increasing energy costs are driving many of Alaska’s more than 200 rural communities to search for alternative sources. Moreover, prices continue to rise as every item in the supply chain must be shipped, barged, or flown into all regions, and prices in the Arctic are often twice those in the contiguous United States (Goldsmith and Rowe 1983).

Alaska’s digital infrastructure is also limited. In rural regions, the scarcity of optical fiber connections and heavy reliance on microwave and satellite connections have resulted in internet connections that are both poor quality and expensive, although the proliferation of Starlink and other low-Earth-orbiting communication devices has begun ameliorating the situation (Barrett et al. 2025). Networks vary in resilience and reliability, and natural disruptions, such as sea ice severing fiber optic lines, exacerbate connectivity issues (Grindal and Meinrath 2025). Additionally, a lack of the utility infrastructure required to build and develop resilient energy systems is compounding cost issues and resulting in the frequent deployment of on-site generation and power storage (Seto et al. 2024).

The population sizes of rural Alaskan settlements range from 13 to 6,182 (Allen et al. 2016). These communities’ limited labor pools necessitate flying in experts with specialized skills. Importing skilled workers inflates these projects’ baseline cost.

Alaska’s limited workforce, high number of grids with low short-circuit capacity, and infrastructural constraints point to the state’s need for more stable and resilient infrastructure.

SMRs and Social License Realities

Alaska’s extreme cold, isolation, and increased need for electricity have drawn attention to advanced nuclear reactor technologies, including SMRs. Notably, the Eielson Air Force Base test-bed program (Air Force 2021) and the U.S. Army’s Janus program at Fort Wainwright (U.S. Army Public Affairs 2025), both in the state’s Interior, serve as potential sites for these technologies. SMRs provide a reliable flow of electricity and thermal energy, and their modularity allows for incremental expansion (Holdmann et al. 2021). Additional units can be integrated into the existing system as needed without requiring a total overhaul of the infrastructure.

To avoid social friction, however, technologies like SMRs and large national security infrastructure need SLO, as the Arctic Over-the-Horizon Radar (A-OTHR) transmitter site project for NORAD in Canada has demonstrated. Successful implementation depends on local stakeholders’ approval and trust. SLO is particularly critical in the North American Arctic because past negative experiences have influenced current perceptions of and responses to new projects.

The federal government and its military’s past actions continue to influence local perceptions and relationships in Alaska. Those actions include proposing to use thermonuclear detonations to excavate a deep harbor (O’Neill 2007), conducting three high-yield underground nuclear tests (Kohlhoff 2024), and unethically using radioactive iodine on local populations for the Arctic Aeromedical Laboratory’s Thyroid Function Study (National Research Council 1996). History teaches us that greater bottom-up engagement with the public yields more durable long-term results (Tavcar 2025).

Data Centers in Alaska

Developments in energy and AI technologies have led to discussions of building data centers in the Arctic. The low ambient temperatures in places like Alaska enable dissipating waste heat via a technique called free cooling, thus limiting the use of chillers and decreasing energy consumption (Fiorillo 2026; Ivanova and Farkhatdinov 2025).

Millions of users now utilize LLMs and other generative AI technologies daily (Berger 2025). To handle AI systems’ data-intensive operations, data center infrastructure needs hardware capable of providing high-reliability bandwidth (generally through fiber optic cables), power efficiency, and low latency (Saber et al. 2026). AI data centers currently demand up to 100 MW of power per location (Colangelo et al. 2025), and their power consumption may grow as the technology develops. Energy is thus a bottleneck for potential AI data centers attached to Alaskan microgrids. Fortunately, dual-use infrastructure provides a potential avenue for achieving energy resilience while both powering data centers and stabilizing Alaska’s microgrids.

Dual Use and Infrastructure Stability

In light of competing interests’ potential to hamper large infrastructure projects, a dual-use approach can serve as a pathway to meeting existing needs and requirements for the future, enhancing energy resilience, and cultivating acceptance. SMRs illustrate how dual-use technologies can serve both defense and civilian interests. A dual-use approach would enable Arctic communities to achieve infrastructure goals that would be financially or logistically impossible without large military spending or capability support (RANE 2025). Moreover, a dual-use project with SLO has the potential to create a resilient system. Local groups would be more likely to remain committed to the project even during setbacks if they share in the responsibility and oversight, including during the preparation, mitigation, response, and recovery phases. Dual-use infrastructure would bolster microgrid communities’ resilience while also facilitating data center development.

Recommendations

1. Critical infrastructure projects, including for SMRs and data centers, should secure SLO from the relevant communities.

2. Military and civilian stakeholders should build energy infrastructure that is resilient and able to meet the power demands of modern technologies, such as AI data centers.

Conclusion

Investing in critical energy infrastructure with dual-use capabilities in Arctic NATO territories, like the U.S. state of Alaska, has never been more crucial. Amid geopolitical tensions with adversaries like Russia and China, the Western allies must bolster their energy capabilities in the North. As they do so, they must holistically integrate projects into communities, civilian and defense. Acquiring SLO will be central in building SMRs and data centers and bolstering resilience. Rather than follow a top-down approach, governments should embed processes for securing ongoing community approval in every phase of any project.

References

ACEP. 2024. “Remote Alaska Communities Benefit from UAF Renewable Energy Research.” ACEP News. May 3, 2024. https://www.uaf.edu/acep/news/2024/remote-alaska-communities-benefit-from-uaf-renewable-energy-research.php.

Air Force. 2021. “Eielson AFB Announced as Site for Air Force Micro-Reactor Pilot.” U.S. Indo-Pacific Command. October 18, 2021. https://www.pacom.mil/Media/News/News-Articles/Article/2814138/eielson-afb-announced-as-site-for-air-force-micro-reactor-pilot/.

Allen, Riley, Donna Brutkoski, David Farnsworth, and Peter Larsen. 2016. Sustainable Energy Solutions for Rural Alaska. Ernest Orlando Lawrence Berkeley National Laboratory. https://doi.org/10.2172/1249184.

Barrett, Will, Sharbel Kozhaya, Zaher M. Kassas, and David Marsh, "Navigating the Arctic Circle with Starlink and OneWeb LEO Satellites," MILCOM 2025 - 2025 IEEE Military Communications Conference (MILCOM), 2025, pp. 1-6, doi: 10.1109/MILCOM64451.2025.11310048.

Berger, Adam. "Artificial Intelligence Data Centers and United States Based Hyperscalers: Impacts and Solutions." (2025) https://jscholarship.library.jhu.edu/handle/1774.2/70474.

Cloudflare. n.d. “What Is a Large Language Model (LLM)?” Cloudflare. Accessed June 4, 2026. https://www.cloudflare.com/learning/ai/what-is-large-language-model/.

Colangelo, Philip, Ayse K. Coskun, Jack Megrue, Ciaran Roberts, Shayan Sengupta, Varun Sivaram, Ethan Tiao et al. 2025. “AI Data Centres as Grid-Interactive Assets.” Nature Energy 11 (2): 254–61. https://doi.org/10.1038/s41560-025-01927-1.

Eschenbach, Ted G., George A. Geistauts, and A. T. Stoddard. 1989. “Alaska’s Infrastructure and Limits to Growth.” Journal of Cold Regions Engineering 3 (1): 23–36. https://ascelibrary.org/doi/epdf/10.1061/%28ASCE%290887-381X%281989%293%3A1%2823%29.

Fiorillo, Luca, Sharon Ridolfo, Attilio Caravelli, and Dario Milone. 2026. “Cooling Strategies for Energy-Efficient Data Centers: Narrative Review and Future Perspective.” International Research Journal of Multidisciplinary Scope 7 (02): 1351–64. https://doi.org/10.47857/irjms.2026.v07i02.08460.

Goldsmith, Oliver Scott, and Phillip Rowe. "The Economic Impacts of Capital Construction Expenditures in Alaska." (1983). Reports .1392. https://ualaska.researchcommons.org/uaa_iser_reports/1392

GRID-Arendal. 2013. Boundaries of the Arctic Council Working Groups. GRID-Arendal. Accessed June 4, 2026. https://www.grida.no/resources/8387.

Grindal, Karl, and Sascha Meinrath. 2025. “Analyzing Broadband Reliability: Insights from Alaskan Arctic Communities.” Journal of Information Policy 15:404–40. https://doi.org/10.5325/jinfopoli.15.2025.0012.

Holdmann, Gwen, George Roe, Steve Colt, Heike Merkel, and Kathy Mayo. 2021. “Small-Scale Nuclear Power: An Option for Alaska?” https://www.uaf.edu/acep/files/working-groups/nuclear-energy/ACEP_Nuclear_Report_2020.pd.

Holdmann, Gwen. 2025. "At the Crossroads of Technology, Policy, and Society: Energy Transitions in Rural Alaska." Order No. 31935313, University of Alaska Fairbanks. https://www.proquest.com/dissertations-theses/at-crossroads-technology-policy-society-energy/docview/3201334671/se-2.

IAEA. 2019. “Maintaining a Social Licence.” IAEA. Accessed June 4, 2026. http://www.iaea.org/resources/nuclear-communicators-toolbox/basics/trust/social-licence.

Ivanova, Mariya, and Nail Farkhatdinov. 2025. “The Arctic Is Becoming a Hot Spot for Data Centers.” Arctida. November 12, 2025. https://arctida.io/en/projects/free-cooling.

Kohlhoff, Dean W. 2024. “Among the Many Islands.” In Amchitka and the Bomb: Nuclear Testing in Alaska, 1–12. University of Washington Press. https://www.degruyterbrill.com/document/doi/10.1515/9780295800509-003/html?lang=en.

Leddy, Laura, Eliza Hotchkiss, and Heidi Applegate. "Conceptualizing Energy Security and Resilience." No. NREL/PO-5R00-89206), in National Renewable Energy Laboratory (NREL), Golden, CO (United States) (2024).https://www.osti.gov/biblio/2344971 

Liou, Joanne. 2023 “What are Small Modular Reactors (SMRs)?” IAEA. September 13, 2023. http://www.iaea.org/newscenter/news/what-are-small-modular-reactors-smrs.

National Research Council. 1996. “Conclusions and Recommendations.” In The Arctic Aeromedical Laboratory’s Thyroid Function Study: A Radiological Risk and Ethical Analysis. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK232521/.

NATO. 2026. “Arctic Security.” NATO. February 16, 2026. https://www.nato.int/en/what-we-do/deterrence-and-defence/arctic-security.

O’Neill, Dan. 2007. The Firecracker Boys: H-Bombs, Inupiat Eskimos, and the Roots of the Environmental Movement. Basic Books.

Peterson, Maren, Magnus de Witt, Ewa Lazarczyk Carlson, and Hlynur Stefánsson. 2026. “Policy and Financial Implications of Net Energy Metering in Arctic Power Systems: A Case Study of Alaska’s Railbelt.” Energies 19 (3): 787. https://doi.org/10.3390/en19030787.

RANE. 2025. The Role of Dual-Use Infrastructure in Securing U.S. Arctic Strategic and Economic Interests. RANE. https://denali.gov/wp-content/uploads/2025/08/ArcticDualUseMay2025.pdf.

Saber, Md Ghulam, and Zhiping Jiang. 2026. “Physical Layer Standardization for AI Data Centers: Challenges, Progress, and Perspectives.” IEEE Network 40 (2): 147–55. https://doi.org/10.1109/MNET.2025.3557812.

Seto, Youji, Hiromi Kageyama, and Pradeeban Kathiravelu. 2024. “Rural Alaska Towards Internet Access and Equity.” In 2024 IEEE Canadian Conference on Electrical and Computer Engineering (CCECE), 910–14. https://doi.org/10.1109/CCECE59415.2024.10667227.

Stanford HAI. n.d. “What is Artificial Intelligence (AI)?” Stanford HAI. Accessed June 4, 2026. https://hai.stanford.edu/ai-definitions/what-is-artificial-intelligence-ai.

Susnjara, Stephanie, and Ian Smalley. n.d. “What Is a Data Center?” IBM. Accessed June 4, 2026. https://www.ibm.com/think/topics/data-centers.

Tavcar, Jus. 2025. “Just Energy Transition to a Low-Carbon Economy in the Arctic: The Case of Alaska.” MA thesis, University of Alaska Fairbanks. Proquest. https://www.proquest.com/pqdtlocal1005781/docview/3285904097/abstract/714B714F3B834733PQ/1.

Ton, Dan T., and Merrill A. Smith. 2012. “The U.S. Department of Energy’s Microgrid Initiative.” Electricity Journal 25 (8): 84–94. https://doi.org/10.1016/j.tej.2012.09.013.

U.S. Army Public Affairs. 2025. “Army Announces Janus Program for Next-Generation Nuclear Energy.” U.S. Army. October 14, 2025. https://www.army.mil/article/288903/army_announces_janus_program_for_next_generation_nuclear_energy.

Funding Acknowledgement

This work relates to the Department of Navy award N00014-22-1-2049 issued by the Office of Naval Research. The Alaska Center for Energy and Power at the University of Alaska provided funding for the SMR analysis.