**Clarifying OpenAI’s Position on Government Guarantees and AI Infrastructure**
There have been some questions and concerns circulating regarding OpenAI’s approach to government guarantees and our infrastructure investments. We’d like to clarify our position and share our thinking on these important topics.
**No Government Guarantees for OpenAI Datacenters**
First and foremost, OpenAI neither has nor wants government guarantees for our datacenters. We believe strongly that governments should not pick winners or losers in the marketplace, and taxpayers should not be asked to bail out companies that make poor business decisions or fail to compete effectively. If one company fails, others will step in and continue to do good work—the ecosystem will be fine.
**Role of Government in AI Infrastructure**
While we are not seeking government support for our own infrastructure, we do see merit in governments building and owning their own AI infrastructure. If governments decide to invest in AI computing power and become major “offtakers” of technology, it may make sense for them to access lower-cost capital to do so. Building a strategic national reserve of computing power—a resource to benefit the government and its citizens, not private companies—makes sense from a policy perspective.
**Semiconductor Fabs and Loan Guarantees**
The only area where we have discussed loan guarantees is in support of US semiconductor fab buildouts. Along with other companies, we have responded to government calls to help strengthen the US chip supply chain for jobs, industrialization, and strategic independence. While we did not formally apply for support, we support efforts to make chip sourcing as American as possible for the benefit of all US companies. Importantly, this is different from governments guaranteeing private datacenter expansion.
**Addressing Key Questions and Concerns**
There are three main “questions behind the question” that understandably cause concern:
**1. How Will OpenAI Pay for Its Infrastructure?**
We expect to end this year with an annualized revenue run rate above $20 billion and aim to grow to hundreds of billions by 2030. Our commitments currently add up to about $1.4 trillion over the next eight years. Meeting these targets will require continued revenue growth and innovative strategies. For instance, we are excited about upcoming enterprise offerings, new consumer devices, robotics, and even categories like AI for scientific discovery. We are also considering offering compute capacity—“AI cloud”—directly to other companies and people. Additional equity or debt raisings are possible. Overall, the demand for computing power is immense, and we’re preparing to meet it.
**2. Is OpenAI Trying to Become “Too Big to Fail”?**
Absolutely not. If OpenAI falters and cannot recover, we should fail—other companies will continue to provide great products and services. That’s how capitalism works, and the economy will adapt accordingly. We are determined to be a successful company, but if we make mistakes, it’s on us.
This week, our CFO discussed government financing and later clarified her remarks, emphasizing the need for a clear national strategy for government-owned AI infrastructure. Separately, I recently spoke with Tyler Cowen about the federal government’s role as a “last-resort insurer” for catastrophic risks, such as rogue actors orchestrating large-scale cyberattacks using AI. To be clear, this does not involve governments bailing out AI companies—it’s about governments managing systemic risks. We do not advocate for the government to write insurance policies for AI companies.
**3. Why Invest So Much Now Instead of Growing Slowly?**
Our vision is to help build the infrastructure for a future economy powered by AI. Based on current research and market trends, now is the right time to invest and scale up our technology. Massive infrastructure projects take years to develop, so early action is essential. Right now, compute constraints force us and others to rate-limit products and delay new features. We believe the risk of not having enough compute capacity is far more significant than the risk of having too much. We want to be ready when AI can drive groundbreaking scientific advances and solutions, perhaps even helping to cure deadly diseases.
We aspire to make AI abundant and affordable, knowing the technology can improve lives in countless ways. Building infrastructure at this scale is a huge privilege and responsibility, and we’re committed to making the right bet for the future. However, if we are wrong, the market—not the government—will address the consequences.
**Conclusion**
We hope this clarifies our stance: OpenAI is committed to responsible growth and investment, without seeking government guarantees for our datacenters or private benefit bailouts. We believe in competing on merit, and we welcome robust public and industry debate about how best to build the future of AI for everyone’s benefit.
https://www.shacknews.com/article/146718/openai-ceo-sam-altman-government-ai
