If OpenAI and Anthropic fail to achieve long-term financial profitability, the United States should nationalize these AI labs to operate them as public research institutions under democratic control.

Key Points

• OpenAI and Anthropic face significant market headwinds, including high training costs, rapid model depreciation, and competition from free, open-source alternatives.
• The authors argue that the current for-profit model is unsustainable, as AI models are increasingly viewed as commodities rather than unique, high-margin assets.
• Nationalization would involve restructuring these companies into public agencies focused on innovation and utility-style compute operations.
• Similar public AI labs and national supercomputing centers are already operational in countries like Switzerland, Singapore, Germany, and Australia.
• Public ownership would prioritize democratic oversight, transparent data usage, and pro-social development goals over the pursuit of shareholder profits.

Why it Matters

Transitioning leading AI labs into public entities could prevent the collapse of critical technology infrastructure if private market valuations prove unsustainable. This shift would align the development of powerful AI models with public interest and democratic values rather than the volatile incentives of venture capital.