Inteligencia Artificial

FERC accelerates connection of large loads for AI and manufacturing

New rules allow data centers and AI factories to connect to the power grid faster, reducing costs and strengthening the system.

June 19, 2026 · 4 min read

cable network

TL;DR: FERC issued an order that facilitates the connection of large electricity consumers, such as AI data centers, through rules that allow self-financing of upgrades, generating their own power, and offering flexibility. This accelerates projects, reduces costs, and strengthens the grid.

What happened?

On March 20, 2025, the U.S. Federal Energy Regulatory Commission (FERC) issued a historic order establishing new rules for the interconnection of large loads to the electric system. The measure directly responds to the growing energy demand from artificial intelligence data centers, semiconductor factories, and advanced manufacturing plants, sectors that have seen an exponential increase in their electricity consumption. According to NVIDIA's official blog, the FERC order is "a major milestone in grid infrastructure" that affects how companies building AI factories, semiconductor manufacturing support systems, and advanced manufacturing facilities can connect to the grid.

The new rules allow large consumers to directly finance their own grid upgrades, bring new electric generation in parallel with their demand, and offer flexibility to reduce or shift their consumption during system stress. This accelerates connection timelines, which could be reduced to as little as 60 days for flexible loads, compared to the years traditional interconnection processes could take. The FERC order follows the directive of U.S. Energy Secretary Chris Wright, who ordered the commission to address large load interconnection.

Why is it important?

The FERC decision addresses a critical bottleneck for the growth of AI and advanced manufacturing in the U.S. Data center electricity demand has surged: according to estimates from the Electric Power Research Institute (EPRI), data centers could consume up to 9% of total U.S. electricity by 2030, up from 2.5% today. Traditional interconnection processes could take between 3 and 5 years, creating uncertainty for multi-billion-dollar investments. By streamlining the process, uncertainty is reduced and return on investment is accelerated.

Additionally, the measure has a direct impact on electricity prices. According to a study by the Lawrence Berkeley National Laboratory, each 10% increase in state electricity consumption is associated with a reduction of approximately 6 cents per kilowatt-hour in retail rates. By adding new demand efficiently, fixed costs are spread over a broader base, which can lower rates for all consumers. NVIDIA highlights that the FERC order is "a pro-growth, pro-affordability, and pro-reliability policy."

Consequences for industry and users

For technology and manufacturing companies, the new regulation means they can bring their facilities online faster, reducing construction timelines for data centers and factories. This is particularly relevant for companies like NVIDIA, which drive the expansion of AI infrastructure. Jensen Huang, CEO of NVIDIA, has described AI as a "five-layer cake," where energy is the fundamental foundation of technological innovation. The FERC order allows large consumers to no longer be "passive entrants in an overloaded interconnection queue" but active participants in building the infrastructure they need.

For residential and commercial consumers, the main potential benefit is stability or reduction in electricity rates, provided the new demand is integrated in an orderly manner. However, there is a risk that if large loads are not flexible, they could increase pressure on the grid during peak hours, potentially requiring additional investments in generation capacity. The FERC order addresses this by incentivizing flexibility: loads that can reduce or shift their consumption during stress periods receive faster connection timelines.

For grid operators, the ability to manage flexible loads offers an additional tool to maintain system stability, reducing the need to build new generation capacity solely for peak demand. This is similar to existing demand response programs, but now on a much larger scale.

What readers should know

  • The FERC order is a significant policy change that aligns the incentives of large consumers with those of the grid. For the first time, large consumers can finance their own grid upgrades and bring new generation in parallel, reducing pressure on existing ratepayers.
  • It is expected to accelerate the construction of AI data centers, especially those that can offer flexibility in their consumption. NVIDIA has noted that "this is not just faster interconnection. It's smarter interconnection," highlighting the role of flexibility.
  • The measure could serve as a model for other countries facing similar challenges in large load interconnection, such as the European Union or China, where data center demand is also growing rapidly.
  • However, implementation will depend on local utilities and the ability of developers to meet the new requirements. The FERC order establishes a national framework, but local utility companies will be responsible for conducting interconnection studies and approving projects.
"This is not just faster interconnection. It's smarter interconnection," NVIDIA noted in its blog, highlighting the role of flexibility in the new regulation.

In summary, the FERC order represents a paradigm shift in how the United States integrates large energy consumers into the grid. By allowing large consumers to finance their own upgrades, bring new generation, and offer flexibility, it accelerates the connection of AI data centers and advanced factories, reduces costs for all consumers, and strengthens grid reliability. However, success will depend on careful implementation and collaboration among utilities, developers, and regulators.

Keep reading