Lately, much of the conversation around AI has centered on the infrastructure necessary to power AI. In particular, the conversation has focused on the need for data centers—more data centers, larger data centers, more advanced data centers. This article will touch upon the evolving data center ecosystem, including the regulatory environment, burgeoning litigation, and practical impediments such as the lack of a robust energy infrastructure.
Massive Investment in Data Centers
The amount of capital being poured into AI can hardly be underestimated. In July 2025, Meta alone stated it plans to “invest hundreds of billions of dollars” into building “superintelligence.”1 A large chunk of that investment is building data centers. Meta plans several AI campuses across the country, all with appropriately supersized names drawn from Greek mythology: Titans like Prometheus and Hyperion. Meta’s competitors, like Anthropic and one of its investors, Amazon, are likewise sinking significant capital into data centers.2
Earlier this year, there was much talk about whether AI might require less computing power than previously thought. News about the Chinese company DeepSeek and its supposed ability to do more with less—that is, produce high-quality AI output at a fraction of the computing power and cost—temporarily cast doubt on projections of the need for so many data centers.
That hesitation appears to have passed, though, and companies are racing ahead to build data centers at an ever-growing pace. As that race continues, the regulatory setting appears to be shifting in favor of data centers.3
Federal Deregulation
In July 2025, the federal government issued “America’s AI Action Plan.”4 This 28-page document lays out three pillars. In addition to accelerating AI innovation and leading in international AI diplomacy and security, one of the pillars is building American AI infrastructure. The very first recommended policy action under this pillar is streamlining the permitting of data centers.
The recommended policy actions include: establishing Categorical Exclusions under the National Environmental Policy Act (NEPA), meaning that environmental assessments and impact statements are unnecessary because the activity does not have a significant effect on the environment; exploring a nationwide Clean Water Act Section 404 permit for data centers that does not require a Pre-Construction Notification; and expediting permits under other acts including the Clean Air Act and the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA, a/k/a Superfund).
In short, the current administration is seeking to deregulate the data center environment to allow for the construction of more, and larger, data centers to fuel the rise of homegrown AI.
Other Constraints Remain
Even if federal regulations are largely minimized, state and local regulations will remain in play. An earlier draft of the “One Big Beautiful Bill Act” would have preempted state and local regulation of AI. However, that provision did not make it into the final bill.5 That means companies may have to navigate local regulatory environments.
Beyond regulations themselves, the threat of litigation also looms for data centers. Some litigation is already underway. For example, Elon Musk’s AI company, xAI, has faced a lawsuit at its data center in Memphis, Tennessee, over claims of air pollution by gas-fired generators.6 Similar environmental-based lawsuits have been filed against companies behind data centers in states from New York7 and Oklahoma8 to Virginia9 and California.10
Besides regulations and potential lawsuits, another issue confronting the construction of data centers is the limited capacity of America’s electrical grid. In the face of overloaded grids, some companies are seeking to generate their own power on-site. But even then, utility companies might not favor such arrangements.11 Like everyone else, utility companies are coming to grips with the rise of AI, including how to meet rocketing energy demands.
What Comes Next?
The titans of the tech industry will continue investing into data centers. With the federal government looking to “cut the red tape” as much as possible, companies building data centers might need to focus on meeting state and local regulations while ensuring that energy needs can be met.
[1] https://www.bisnow.com/national/news/data-center/meta-to-spend-hundred-of-billions-of-dollars-building-manhattan-sized-data-center-campuses-130140?utm_source=outbound_pub_23&utm_campaign=outbound_issue_87496&utm_content=outbound_link_2&utm_medium=email
[2] https://time.com/7273288/amazon-anthropic-openai-microsoft-stargate-datacenters/
[3] https://www.spglobal.com/market-intelligence/en/news-insights/research/potential-impacts-of-deepseek-on-datacenters-and-energy-demand
[4] https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf
[5] https://www.nlc.org/article/2025/07/11/local-impacts-from-congress-one-big-beautiful-bill/
[6] https://www.newsweek.com/elon-musks-ai-data-center-faces-lawsuit-over-air-pollution-complaints-2087032.
[7] https://www.wxxinews.org/local-news/2025-07-02/lawsuit-targets-data-center-proposed-for-genesee-county-industrial-park
[8] https://www.youtube.com/watch?v=v_AzE0t2RPI
[9] https://www.youtube.com/watch?v=3UgKl18Iyjo
[10] https://biologicaldiversity.org/programs/urban/pdfs/Verified-Petition-for-Writ-of-Mandate-and-Complaint-Final-w-exhibits.pdf
[11] https://www.bisnow.com/national/news/data-center/utilities-slowing-development-of-self-powered-data-centers-130127?utm_source=outbound_pub_23&utm_campaign=outbound_issue_87496&utm_content=outbound_link_13&utm_medium=email