Biden orders federal lands opened for gigawatt-scale AI data centers
President Biden just signed an executive order to fast-track US AI data center construction by leasing federal lands. The Defense and Energy departments must identify at least three sites each by February 28th to host gigawatt-scale facilities. While developers must fund clean power and source US-made chips, the move sparks heavy backlash over grid strain and climate goals.

The race for AI dominance has officially run out of conventional real estate.
President Biden just signed an executive order aimed at drastically speeding up the construction of AI data centers across the US. The directive makes federal land available to tech giants and tells agencies to clear the bureaucratic hurdles slowing down the infrastructure buildout.
Federal lands and fast-tracked permits
The core of the order directs the Department of Defense and the Department of Energy to lease federal land to private companies building gigawatt-scale AI data centers alongside clean energy facilities. Both departments must identify at least three suitable sites each on the lands they manage by February 28th.
Speed is the name of the game here. Federal agencies are instructed to prioritize AI infrastructure permits. The measure could also introduce categorical exclusions under the National Environmental Protection Act, allowing developers to sidestep prolonged environmental reviews.
To counter public backlash, the White House is putting the financial burden squarely on developers. Companies must pay all costs for the facilities, new power plants, and transmission lines so consumers do not see their utility bills spike. On top of that, operators are required to source power from new clean energy projects, evaluate the national security risks of models trained on federal land, and purchase an appropriate share of American-made chips.
The looming clash over power and climate
The White House is moving aggressively because the numbers are alarming. According to estimates published on December 20th by the Lawrence Berkeley National Laboratory, data center power demand has tripled over the past decade. It could double or triple again by 2028, potentially jumping from 4.4 percent of total US electricity consumption in 2023 to as much as 12 percent.
Utilities across the country are already delaying the retirement of dirty coal and gas plants just to keep the lights on for power-hungry servers. That reality has sparked fierce pushback.
On December 17th, five Democratic senators, including Elizabeth Warren and Sheldon Whitehouse, warned the administration against bypassing environmental safeguards. Johanna Neumann from Environment America Research & Policy Center similarly cautioned that without strict guardrails, AI's insatiable thirst for power risks crashing efforts to transition away from fossil fuels.
Why it matters
Power and land have officially overtaken chips as the ultimate bottleneck for frontier AI. We think this order is a shrewd move to bypass local zoning gridlock by tapping directly into federal reserves, signaling that the US treats data center capacity as critical national defense infrastructure.
For major cloud providers and foundation model developers, this could unlock gigawatt-scale projects that were previously bogged down in local red tape. But the promises made to ordinary consumers look much harder to keep.
Even if tech companies pay for their own clean generators, they still rely on the broader grid infrastructure. As AI clusters demand unprecedented amounts of juice, keeping aging fossil fuel plants alive to maintain baseload stability seems almost inevitable. Balancing the AI race with climate reality remains a paradox that executive pen strokes alone cannot easily resolve.
Sources (1)
- theverge.com(opens in a new tab)Originaltheverge.com