Free tools Windows power users keep installed
One-click scans. No signup required.
OpenAI pitched the Biden administration on policies that could enable enormous U.S. AI data centers, with proposed facilities potentially requiring 5 gigawatts of power apiece. That is roughly comparable to the output of five nuclear reactors or the electricity used by nearly three million homes. The September 2024 proposal was not an approved construction network, a federal funding commitment, or a Biden-backed OpenAI project.
What OpenAI proposed
In a document shared with U.S. officials, OpenAI discussed building very large data centers in various states, potentially starting with one facility and expanding. The reported power requirement was as much as 5 gigawatts for each center. OpenAI’s document did not settle how many centers would ultimately be built or specify a final network of locations. Bloomberg Law reported the proposal on September 25, 2024.
![AI Law and Policy: [Connected eBook]](https://m.media-amazon.com/images/I/41Y7jOay2eL._SL160_.jpg)
Constellation Energy CEO Joe Dominguez said he had heard Altman was considering five to seven facilities. That was Dominguez recounting what he had heard, not a confirmed OpenAI commitment.
Why OpenAI wanted federal support
OpenAI framed expanded AI infrastructure as a way to preserve U.S. leadership, compete with China, support domestic industry and jobs, and make computing capacity available for future AI development. The company’s spokesperson described the infrastructure as important to innovation, reindustrialization, and access to AI’s benefits. Those are OpenAI’s arguments, not independently verified forecasts of economic or national-security outcomes.
The reported pitch was for supportive conditions that could make construction possible—not a disclosed federal appropriation to build OpenAI’s facilities. In practice, “support” can cover very different things, from coordinating permits and grid planning to providing public land or incentives. The reporting establishes a policy pitch, but not that Biden approved any particular subsidy, loan guarantee, or financing arrangement.
What happened at the White House
On September 12, 2024, the White House held a broader roundtable on U.S. AI infrastructure. Attendees included Altman, Nvidia CEO Jensen Huang, Anthropic CEO Dario Amodei, Google President Ruth Porat, Amazon cloud chief Matt Garman, Microsoft President Brad Smith, energy-sector representatives, and senior Biden officials. The discussion addressed electricity generation, grid capacity, data-center construction, workforce, permitting, semiconductor supply, and national security; it was not an OpenAI-only presentation. The White House readout announced a federal Task Force on AI Datacenter Infrastructure and a Department of Energy engagement team.
Rank #2
What a 5-gigawatt requirement means
Five gigawatts is an extraordinary power scale. Bloomberg Law compared it with the output of roughly five nuclear reactors and electricity consumption equivalent to nearly three million homes. These are rough comparisons, not a plan to build five reactors for each data center.
A gigawatt figure also needs context. It may refer to a target or available power capacity, a facility’s peak demand, or a sustained load; those are not interchangeable. The data center’s IT equipment is only part of the total facility demand, which also includes cooling and other systems. And supplying a stated capacity reliably can require more generation and grid infrastructure than the facility’s typical average use. The reported proposal does not establish a precise operating load profile for each center.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsRank #3
Why building at this scale would be difficult
A data center is not just a large building when its power needs reach this scale. The power supply, grid connections, cooling, equipment, and approvals can be as consequential as the real estate. Energy executives questioned whether even one 5-gigawatt facility could be supplied reliably, according to Bloomberg Law.
- Grid connections and transmission: Interconnection queues and new transmission lines can delay access to power, even where generation exists.
- Generation and reliability: New power plants or other supply arrangements take planning and approvals; a large facility also needs dependable service when demand is high.
- Permits and equipment: Federal, state, and local approvals, as well as shortages of transformers and other equipment, can constrain construction timelines.
- Workers, chips, and cooling: Skilled labor and semiconductors are essential inputs. Cooling systems also bring water and energy demands.
- Local costs and acceptance: Land use, water use, emissions, and the possibility that grid upgrades affect other electricity customers can prompt public concern or opposition.
What the Biden administration did—and did not do
The administration’s response was to coordinate policy and address infrastructure needs across the sector. The task force and DOE engagement team followed the September roundtable. A later Biden executive order addressed AI data centers’ energy requirements, including use of federal sites and support for clean-power infrastructure; it also raised issues involving grid effects and electricity prices. The Associated Press reported on that action.
These were broader federal measures, not evidence that the government approved or funded OpenAI’s particular 5-gigawatt proposal. The September 2024 reporting does not establish a direct taxpayer-funded bailout, federal financing guarantee, or authorization to construct a defined OpenAI network.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The trade-offs behind the national-security case
OpenAI’s case links domestic computing capacity to U.S. competitiveness and security. That argument reflects a real policy question: whether the country can support strategically important AI infrastructure while managing grid reliability, affordability, and environmental effects. It also serves OpenAI’s commercial interest in obtaining more computing capacity.
- Potential public benefits OpenAI cited: domestic industrial activity, jobs, faster access to compute, and a stronger U.S. position in AI.
- Public concerns: pressure on local power systems, electricity prices, emissions and water use, and whether public support would disproportionately benefit a few large technology companies.
- Forecast risk: investments sized around expected future demand could become costly if demand does not materialize as projected.
OpenAI’s claims about jobs, GDP, and national security should therefore be read as advocacy arguments, not guaranteed results. The policy choice is not simply whether AI matters; it is who bears the infrastructure costs and how benefits are distributed.
How the 2024 pitch relates to later infrastructure plans
In January 2025, OpenAI, SoftBank, Oracle, and partners announced Stargate, a separate infrastructure initiative publicly associated with a $500 billion, 10-gigawatt commitment. Later announcements described additional sites and capacity under that initiative. OpenAI’s announcement of five new Stargate sites is a later development, not proof that the Biden administration approved the 2024 pitch.
In September 2025, OpenAI and Nvidia announced a partnership targeting at least 10 gigawatts of Nvidia systems for OpenAI infrastructure, with Nvidia intending to invest up to $100 billion as systems were deployed. OpenAI described the partnership as a separate corporate arrangement. Neither later announcement turns the earlier Biden-era proposal into a government-funded project.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
