Sam Altman has revived his 2021 proposal for a national AI wealth fund, now reportedly discussing a 5% equity stake in OpenAI with the Trump administration. The concept draws from Bernie Sanders' more aggressive 50% stake proposal and the Alaska Permanent Fund model, which distributes oil revenues to residents. Altman argues that since AI models train on human-generated content — books, art, code — without compensating creators, a public equity stake serves as belated payment while buffering potential labor displacement.

OpenAI's March funding round valued the company at $852 billion, making a 5% stake worth $42.6 billion. Distributed across 133 million U.S. households, that equals roughly $320 per household in raw equity. In practice, the government would likely manage the stake as a sovereign wealth fund, paying out investment returns rather than liquidating shares — contingent on OpenAI ever achieving sustainable profitability, which it has not yet done despite heavy data center spending.

The political calculus is transparent: the Trump administration has pursued equity positions in Intel and revenue shares from Nvidia's China sales. For OpenAI, maintaining White House favor could prevent models from being labeled supply-chain risks and secure help blocking Chinese rivals. Altman pitched the idea shortly after Trump's inauguration, yet five years of discussion have produced no legislative framework.

The proposal functions primarily as narrative engineering. By framing AI as a wealth generator large enough to share, Altman attempts to shift public sentiment — currently majority-negative on AI trust and data center construction — while conceding that the technology consumes collective creative output. Whether checks ever materialize is secondary to convincing stakeholders the boom justifies the extraction.