New York freezes 50 megawatt licenses, blocking Bitcoin miners' path to AI

According to Woofun AI, a sudden change in New York State's regulatory policy directly froze the relevant power generation permits. This move accurately cut off the key path for Bitcoin mining companies to transform into the field of artificial intelligence. Industry adjustment plans originally aimed at dealing with cryptocurrency fluctuations came to a standstill as the state extended its regulatory reach from a single mining facility to a wide range of areas covering cloud services and large-scale computing projects.

Although the order did not directly ban Bitcoin mining, its target infrastructure characteristics closely coincided with data centers that mining companies urgently needed to be renovated. As a result, companies with strong electricity supplies and industrial land were unable to transform their facilities into AI centers to accommodate graphics processors as planned, and the transformation window was forcibly closed.

Economic pressure within the industry is forcing mining companies to find a way out, but the actual resistance far exceeds expectations. The Financial Times revealed that Keel originally planned to consolidate 96 megawatts of electricity scattered across three Bitcoin mining facilities into a single artificial intelligence data center to build a high-performance computing business.

However, as the market environment deteriorated, the financial viability of the transformation was seriously challenged. According to CoinShares data, in the fourth quarter of 2025, the average cash cost of each Bitcoin produced by listed Bitcoin mining companies soared to about $79,995. At the same time, revenue per unit of computing power fell to a multi-year low.

This scissor gap has made the transformation of AI, which was originally a safe haven, extremely difficult. Companies not only have to bear high transformation costs, but also face the risk of capital chain breakdowns due to project delays. Companies such as Keel have to re-evaluate their operating strategies before the Bitcoin project is profitable or the site has redevelopment value.

Data compiled by Woofun AI shows that resistance at the political level is shifting from fermenting public opinion to a legislative game, and public concerns about resource consumption have become the core driving force behind multi-state legislative proposals. Of those who opposed local data center projects, more than half clearly indicated excessive consumption of electricity, water, and other resources, in addition to deep concerns about pollution, rising utility bills, and traffic congestion. Pennsylvania lawmakers proposed suspending related projects for three years to conduct environmental impact assessments; the South Carolina bill proposed not granting construction permits until a statewide regulatory framework was established; and Vermont lawmakers proposed limiting the development of new projects until 2030. These proposals all require the federal government to first introduce measures to protect the rights and interests of electricity users, workers, and the environment; otherwise, the construction and expansion of AI data centers will be suspended.

Although the governor of Maine vetoed an 18-month ban, and proposals from Minnesota, New Hampshire, Oklahoma, and South Dakota did not pass, the regulatory model promoted by New York Governor Hoochul has broken the deadlock, providing other regions with a viable model for limiting project development while studying electricity costs and infrastructure needs.

The threshold for future transformation will rise significantly due to the AI industry's stringent requirements for high reliability, and the cost structure is facing a complete restructuring. Unlike Bitcoin mining, which can be suspended when electricity prices soar or the power grid is overloaded, artificial intelligence projects require nearly uninterrupted power supply and strict backup power generation facility guarantees. This makes companies unable to complete the transformation on schedule only continue to rely on volatile Bitcoin production, causing capital to be locked in unfinished projects for a long time. The New York State order revealed the specific form of this cost pressure ahead of time: Hochul directed the establishment of a “grid acceleration fund,” requiring developers to prepay funds to support grid upgrades, clean energy production, and battery energy storage, and take remedial measures for undersized projects.

At the same time, the “beneficiary payment” system will force large customers to bear the burden of power grids and infrastructure, and regulators may also set different electricity service levels, forcing data centers to invest their own funds to build dedicated power generation or energy storage facilities. These measures mean that even if companies have land and substation resources, they cannot avoid high additional costs with grid connection rights alone, which ultimately leads to lengthening project development cycles, a surge in infrastructure investment costs, and a drastic reduction in site selection options suitable for large-scale AI data centers.