OpenAI has held early-stage discussions with the Trump administration about transferring a 5 percent equity stake to the US government, a move that would give the state a direct financial interest in the company behind ChatGPT. The talks, first reported by the Financial Times and confirmed by multiple international outlets, come as Washington tightens oversight of frontier AI models and debates how to distribute the technology’s expected economic windfall. OpenAI, valued at $852 billion after a record funding round in March, has suggested that other leading US AI developers—including Anthropic, Google, and Meta—cede similar stakes, though it remains unclear whether those firms would agree.

The proposed mechanism draws on the Alaska Permanent Fund, a state-owned corporation that invests oil revenues and pays annual dividends to residents. OpenAI’s chief executive, Sam Altman, and other executives have argued that channelling equity into a comparable sovereign wealth vehicle would allow all Americans to share in AI-generated wealth, regardless of their participation in financial markets. The idea echoes earlier policy papers from OpenAI and Anthropic that called for public wealth funds or “digital dividends” funded by sector taxes. Viewed from Washington, the concept has gained traction across the political spectrum: President Donald Trump has described a state stake in AI firms as a “beautiful thing,” while Senator Bernie Sanders has introduced legislation that would require up to 50 percent public ownership of major AI companies through a sovereign fund.

The equity discussions unfold against a backdrop of escalating government intervention. In recent weeks, the Commerce Department ordered Anthropic to suspend foreign access to its Fable 5 and Mythos 5 models over national security concerns—restrictions lifted only on 1 July—and the White House asked OpenAI to delay the full public release of GPT-5.6. Both companies have confidentially filed for initial public offerings, a step that would broaden their shareholder base and generate substantial returns for early investors. Analysts in London note that granting the government a stake could help secure political goodwill and pre-empt more aggressive regulatory or tax measures, particularly as public anxiety grows over data-centre construction, job displacement, and cybersecurity risks.

Any equity transfer would likely require congressional approval, and the talks remain at what participants describe as a conceptual stage. The White House and OpenAI have declined to comment. The next milestone to watch is whether the administration formalises the proposal in a legislative framework or incorporates it into broader AI governance negotiations, which are expected to intensify ahead of the autumn midterm elections.