Voters in communities across the US are backing limits on new data centers, potentially giving Bitcoin miners with existing power capacity and sites an advantage as they pursue AI infrastructure contracts. But converting a mining facility takes more than a change of use: cooling, networking and power reliability all require investment, while rising data center demand is sharpening debate over who pays for grid upgrades.
40 of 41 local votes favored development limits
On November 3, during the US midterm elections, voters in several locations considered data center measures alongside congressional races. Of 41 local ballots tracked by Ballotpedia, 40 favored restrictions. Concerns included higher electricity bills, water use and the effects of facilities on nearby communities.
The measures took different forms, including citywide bans, limits on large facilities, pauses on new projects and requirements for voter approval of certain developments. Ballotpedia said 27 measures restricting or opposing new data centers passed, including five outright bans. Three towns in Maine also approved six-month moratoriums on new projects.
In Grove City, Ohio, voters approved a rule requiring public approval and other conditions for large projects using more than 20 megawatts of electricity. Local control over zoning and building permits means that even companies with a site and power supply may face restrictions on expansion or infrastructure upgrades.
Jordan Knecht, head of strategic integration at Kresus, said local restrictions could make data center projects harder to advance. For miners looking to enter AI, understanding local approvals, power access and development requirements may be as important as tracking federal policy.
AI contracts are growing, but conversion costs are steep
Bitcoin miners are already moving into AI infrastructure. In its first-quarter 2026 report, CoinShares said publicly listed miners had announced more than $70 billion in AI and high-performance computing contracts from 2025 through early 2026.
Existing power access is only a starting point. A 2026 CoinShares report estimated that building an AI-ready facility costs about $8 million to $15 million per megawatt, compared with roughly $700,000 to $1 million per megawatt for a conventional mining facility. The gap reflects AI data centers’ more demanding requirements for cooling, networking and operational reliability—and the substantial capital needed to meet them.
Jamie Brown, CEO of Global Kiwi Labs, which hosts servers for crypto exchanges and trading firms, said local development restrictions could make conversion more attractive to miners that have yet to repurpose their power contracts and facilities for AI. Brown said AI may offer better returns than mining, but whether a project is worthwhile depends on upgrade costs and the revenue it generates.
Yuliya Barabash, founder of SBSB Fintech Lawyers, noted that mining and AI data centers use power differently. Bitcoin miners can typically reduce consumption when electricity prices rise, while AI data centers generally require a steady, continuous supply. Higher power costs or additional regulatory requirements could reduce the profitability of some projects or delay their launch.
Rising data center demand puts grid costs in focus
The International Energy Agency estimates that data centers used 224 terawatt-hours of electricity in 2025. That figure could nearly double to 426 terawatt-hours by 2030. The projected increase is bringing grid expansion and the allocation of related costs into policy discussions.
An incoming Democratic-controlled Congress could scrutinize how large data centers are approved and how their power needs affect consumer costs. Knecht said lawmakers may press major operators to pay a fairer share of infrastructure costs rather than shifting them to ordinary customers. Barabash also expects attention on whether technology companies should cover the cost of new power infrastructure needed to serve their facilities.
Tighter scrutiny would not necessarily mean Congress will block data center construction. Knecht said AI infrastructure is becoming more important to the US economy, sustaining support for investment in domestic computing capacity. For miners, the central calculation remains whether AI revenue can cover upgrades to cooling, networking and power systems: existing power resources may offer a head start, but do not by themselves determine whether a conversion will be profitable.