I would like to clarify a few important points regarding OpenAI’s relationship with government support and our infrastructure plans.
**No Government Guarantees for OpenAI Datacenters**
First and foremost, we do not have — nor do we want — government guarantees for OpenAI datacenters. We believe that governments should not pick winners or losers, and taxpayers should not bail out companies that make poor business decisions or otherwise fail in the market. If one company fails, others will continue to do good work.
Instead, what might make sense is governments building and owning their own AI infrastructure. In such cases, the benefits should accrue to the government itself. We can envision a scenario where governments decide to purchase large amounts of computing power and determine how to use it. To facilitate this, it might be reasonable to provide governments with access to lower-cost capital. Building a strategic national reserve of computing power is a sensible idea — but it should serve the government’s interests, not those of private companies.
**Supporting Semiconductor Manufacturing in the U.S.**
One area where we have discussed loan guarantees is in supporting the buildout of semiconductor fabrication plants (fabs) in the United States. We, along with other companies, have responded to the government’s call to help strengthen the domestic chip supply chain, though we did not formally apply for any loan guarantees ourselves.
The goal here is to ensure that the sourcing of critical chips is as American as possible — bringing back jobs and industrialization to the U.S., and enhancing the country’s strategic position with an independent supply chain. This benefits all American companies and differs fundamentally from governments guaranteeing datacenter buildouts that provide private benefit.
**Addressing Common Questions and Concerns**
There are at least three key “questions behind the question” that are understandably raising concerns:
**1. How is OpenAI going to pay for all this infrastructure?**
We expect to finish this year with an annualized revenue run rate above $20 billion, and we project growth to hundreds of billions by 2030. Over the next eight years, we are looking at commitments of around $1.4 trillion to build the necessary infrastructure.
Obviously, achieving this scale requires continued revenue growth—each doubling is a significant challenge. But we feel optimistic about our prospects, especially with upcoming enterprise offerings and emerging categories like consumer devices and robotics. There are also new, hard-to-predict areas such as AI-driven scientific discovery that we expect to be important.
Additionally, we are exploring ways to sell compute capacity directly to other companies and individuals. We believe the world will need immense amounts of “AI cloud,” and we are excited to meet that demand. While we may raise additional equity or debt capital in the future, current forecasts suggest global computing power demands far exceed even our ambitious plans.
**2. Is OpenAI trying to become “too big to fail,” and should the government pick winners and losers?**
Our answer is a clear no. If we make mistakes and cannot fix them, we should fail, and the market ecosystem will persist with other companies continuing to serve customers and innovate. That is how capitalism works.
Our CFO recently spoke about government financing and later clarified her remarks to emphasize this point more clearly. We do believe the U.S. government should have a national strategy for AI infrastructure, but that is entirely separate from bailing out companies.
For example, Tyler Cowen recently asked if the federal government might become an insurer of last resort for AI risks—similar to how it manages nuclear risk, not supporting overbuild. I responded that, yes, the government will likely serve as insurer of last resort in the sense of managing catastrophic risks, such as a rogue actor using AI for a large-scale cyberattack that disrupts critical infrastructure.
But this is a very different conversation than government-backed insurance policies for AI companies or datacenter financing. The key concern here is intentional misuse of AI causing harm on a scale only the government could address—not corporate financial risk.
**3. Why does OpenAI need to spend so much now, instead of growing more slowly?**
We are building the infrastructure for a future economy powered by AI. Given the breakthroughs we see on the horizon through our research, now is the time to invest aggressively to scale our technology.
Massive infrastructure projects take years to complete. Based on current trends in AI usage, we believe the risk of having insufficient computing power is greater and more imminent than the risk of having too much.
Even today, we and others must rate limit our products and restrict new features because of severe compute constraints. In a future where AI can enable major scientific breakthroughs—at the cost of tremendous computing resources—we want to be ready.
This moment is no longer distant. Our mission compels us to act swiftly to apply AI to complex problems, such as curing deadly diseases, and to deliver the benefits of artificial general intelligence (AGI) as soon as possible.
Moreover, we envision a future of abundant, affordable AI. Massive demand for AI technology is coming, and it will improve people’s lives in countless ways. It is a great privilege to be in this arena, and we are committed to building infrastructure at scale for something so significant.
This is the bet we are making. And while we feel confident given our vantage point, we acknowledge we could be wrong—if so, the market, not the government, will determine the outcome.
—
We hope this clarifies our position on government involvement and the scale of our infrastructure plans. Our goal remains to responsibly lead the AI revolution while maintaining a commitment to market principles and long-term societal benefit.
https://www.shacknews.com/article/146718/openai-ceo-sam-altman-government-ai
