Restricting data centers is a strategic mistake as AI demand surges.

icon币界网
Share
AI summary iconSummary
BitJie warns that restricting data centers could backfire as AI demand grows. Inflation data shows rising energy costs, while on-chain data reveals increasing demands for crypto infrastructure. Ireland’s data centers consume 24% of the nation’s power, and New York has halted new permits until 2026. The article argues that expanding grid capacity is preferable to limiting growth. The U.S. and China are advancing AI infrastructure, while Europe lags due to slow construction and reliance on U.S. cloud services. Power delivery bottlenecks include transformer delays and surging demand for AI hardware.
CoinDesk reports:

In Ireland, data centers consume nearly a quarter of the country’s electricity. New grid connections around Dublin have effectively been frozen for three years. Under revised rules, new projects must provide generation or energy storage capacity on-site or nearby. Amsterdam previously imposed a moratorium on new facilities, while Frankfurt manages site selection through zoning regulations.

In July 2026, New York State introduced a maximum one-year pause on certain environmental permits for new large data centers. State officials stated that this pause would allow them to assess the impact of these facilities on the grid and the environment. As of May 2026, data center projects representing nearly 12 gigawatts of potential demand were queued for grid interconnection in the state.

When the power grid cannot keep up, should the government limit data centers? I believe the government should expand power supply and grid capacity.

As a co-founder and general partner of Roosh Investment Group, I have been investing in technology companies for over five years. Our portfolio includes AI infrastructure, and energy has now become a key focus. This experience has shown me how access to energy infrastructure influences investment decisions—reliable electricity is increasingly becoming a decisive factor in whether a company can scale and attract capital.

As businesses increasingly rely on AI, control over their models and infrastructure has become more critical. Foreign suppliers could impact entire industries if they decide to alter terms, raise prices, or restrict access. Domestic computing power gives countries greater control over the infrastructure that businesses depend on.

Control over AI is becoming a source of economic power.

This helps illustrate the intensity of the competition between the U.S. and China in the AI sector. U.S. companies hold a leading position across the entire AI value chain, from advanced models and chips to large-scale cloud platforms. Washington treats the supply of data center infrastructure as a national security issue.

Beijing is advancing a state-led strategy. Its "East Data, West Computing" initiative directs computing infrastructure development to western China, where energy and land resources are more abundant. Meanwhile, China is also significantly expanding its power generation capacity. According to the International Energy Agency, China added nearly 500 gigawatts of new renewable energy capacity in 2025, accounting for more than 60% of global growth.

Meanwhile, Europe may fall further behind. Brussels plans to build up to seven AI superfacilities and at least triple the EU’s data center capacity within the next five to seven years. However, some countries are slowing construction, and European businesses still rely primarily on U.S. cloud providers. Restrictions will not eliminate demand for AI; they may redirect investment overseas and deepen this dependency.

The bottleneck lies locally.

The power demand of data centers should be understood within the broader context of overall electricity consumption growth. According to the International Energy Agency, global electricity consumption will increase by 800 terawatt-hours in 2025, with data centers accounting for only about 70 terawatt-hours. Industrial use, air conditioning, heat pumps, and electric vehicles contribute significantly more.

The real challenge is that rapidly growing demand is concentrated in just a few locations. Electricity must be delivered in time to new computing facilities.

What is slowing down the connection?

In the United States, delivery times for large transformers have now exceeded two years, compared to about 50 weeks before the pandemic. Morgan Stanley estimates that by 2028, U.S. data centers could face a power shortfall of approximately 49 gigawatts. Microsoft CEO Satya Nadella acknowledged that power has become a hard constraint for the company—some of its GPUs remain in inventory with nowhere to plug in.

Building dedicated gas-fired power plants is not a quick solution. GE Vernova, Siemens Energy, and Mitsubishi Power produce about two-thirds of the world’s gas turbines, and their factory orders are already booked years ahead. GE Vernova reported that its orders and capacity reservations for gas power equipment have reached 116 gigawatts, with a goal to increase annual turbine production to 20 gigawatts by the third quarter of 2026.

The hardware itself introduces another challenge. Previously, a single server rack required only 7 to 12 kilowatts. Systems built around the NVIDIA GB200 consume 120 to 132 kilowatts per rack. At this power density, traditional air cooling is no longer sufficient, so new facilities must be designed for liquid cooling from the outset.

Tech giants are also securing power directly. Microsoft has agreed to purchase electricity from the 835-megawatt Three Mile Island reactor, which is being restarted to meet its demand. Google’s agreement with Kairos Power involves small modular reactors, with the first expected to come online around 2030.

Governments cannot rely solely on corporate self-initiative. They must plan energy infrastructure alongside digital infrastructure: expanding the power grid, accelerating transformer production, streamlining connection processes, and encouraging long-term power purchase agreements, on-site generation near data centers, and energy storage.

How war accelerated Ukraine's distributed generation

Ukraine’s experience demonstrates how crises drive the deployment of local power generation. War and repeated attacks have forced the country to rapidly deploy local power sources, as the centralized grid can no longer ensure reliable electricity supply.

My experience investing in technology companies has shown that reliable infrastructure is critical for business growth. In Ukraine, businesses have long had to manage energy risks—they need stable power, backup power, and contingency plans for potential outages. This is equally important for investors, as a company’s ability to maintain operations during power outages influences investors’ assessment of its growth potential.

Ukraine currently has approximately 1.8 gigawatts of distributed gas-fired generation capacity installed, with over 500 megawatts under construction. The country is also streamlining grid connection and permitting procedures.

Ukraine’s experience cannot be directly applied to large AI data centers. However, it demonstrates that prioritizing energy resilience can accelerate new power generation investments and drive reforms in connection procedures, even under extreme conditions.

How to power the expansion of a data center

Powering the new data center requires coordinated action from the government, grid operators, and businesses. The government can identify and plan suitable locations for new facilities. Grid operators can reserve capacity in advance. Businesses can invest in their own power plants, energy storage, and cooling systems, and reduce electricity usage during peak hours in exchange for faster grid access. A combination of grid power, natural gas generation, nuclear power, and renewable energy helps supplement supply in areas where no single source is sufficient to meet demand.

The ability to rapidly build and power data centers may increasingly become a key factor in investment decisions. Advantage in the AI race depends not only on access to models and chips, but also on a country’s capacity to simultaneously expand computing power, electricity generation, and the grid. Restricting data center development may temporarily ease pressure on the power grid, but could also deepen dependence on foreign infrastructure and technology.

The views expressed in the article on Fortune.com are solely those of the author and do not necessarily reflect the views and positions of Fortune.

Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.