The Financial Times reported on July 2, 2026 that Sam Altman had proposed transferring approximately 5 percent of OpenAI’s equity to the U.S. federal government. Two people familiar with the discussions confirmed the size to the FT, which described the talks as still “conceptual” but active with President Trump personally, Treasury Secretary Scott Bessent, and Commerce Secretary Howard Lutnick. Bloomberg, CNBC, CNN, and Forbes all subsequently picked up the story with the same core facts and the same official response — that is, no official response.

The story matters because OpenAI is not paying its way in. The Intel deal in August 2025 involved the government making an $8.9 billion investment in exchange for 9.9 percent of the chipmaker, structured through CHIPS Act grants that Intel had been awarded but not yet received. What Altman put on the table is equity without receiving cash in return. Trump himself framed it directly in his August 2025 Truth Social post announcing the Intel deal: “The U.S. paid nothing for these Shares, and the Shares are now valued at approximately $11 Billion Dollars.” That model — government gets equity, company gets political peace — is what Altman is now trying to formalize.

That is a structurally different transaction from a normal capital raise, and it is why the political framing quickly shifted from “government investment” to “corporate donation.” The FT’s original headline captured the direction: OpenAI “proposed” the stake. It was not asked for. Altman offered it.

Altman and other OpenAI executives specifically pointed to the Alaska Permanent Fund. That is a state-owned corporation Alaska established in 1976 to hold a share of its oil royalties. It now manages roughly $80 billion in assets and pays every resident an annual dividend of between $1,000 and $3,000, depending on investment returns. It is one of the only functioning sovereign wealth funds inside the United States, and it is politically popular across ideological lines in a way that federal programs rarely are.

The pitch, then, is not just “give the government a stake.” It is “seed a federal Alaska fund with the equity that America’s largest AI companies are willing to donate, and let the returns flow to citizens.” OpenAI’s April 2026 policy paper laid this out explicitly under the label “Public Wealth Fund”, arguing that every citizen — including those who never invested in the market — should have a stake in AI-driven growth. In May 2026, the OpenAI Foundation, the nonprofit arm, published a follow-up arguing that “in an AI-led future, society will likely need new approaches that give people durable stakes in the systems creating value.”

Trump’s own February 2026 executive order created the legal wrapper for this. It called for the establishment of a federal sovereign wealth fund and directed Treasury and Commerce to develop a plan within 90 days. The order did not specify how the fund would be seeded. Altman’s proposal answers that question with a specific mechanism: AI equity, delivered voluntarily by the industry, starting with 5 percent of OpenAI.

The Trump administration has been building this playbook for a year. In August 2025, the government took a 9.9 percent stake in Intel, worth $11.1 billion in total. The structure was unusual: $5.7 billion came from CHIPS Act grants Intel had been awarded but not yet paid, $3.2 billion from Defense Department Secure Enclave money, plus $2.2 billion Intel had already received. The government bought 433.3 million primary shares at $20.47 per share and took no board seat. Intel confirmed the arrangement in a statement citing “the confidence the Administration has in Intel to advance key national priorities.”

The Nvidia and AMD deal, earlier in August 2025, was a different structure: both companies agreed to pay the government 15 percent of their China chip sales in exchange for H20 export licenses, later reportedly bumped to 25 percent on the H200. Nvidia declined to comment on the specific percentage to NPR but said in a statement it follows “rules the U.S. government sets for our participation in worldwide markets.” Trump also secured a “golden share” veto in the Nippon Steel merger with U.S. Steel, giving the government the ability to block certain company decisions.

Treasury Secretary Bessent has said the model may expand to other industries, and Trump has publicly floated similar arrangements with defense contractor Lockheed Martin. The Intel deal was politically contested — some Republicans, including Senator Rand Paul, called it a departure from free-market principles — but it survived. That survival is what convinced Altman the concept was viable, and it is why he moved from private conversations in early 2025 to a documented FT-sourced proposal in July 2026.

Senator Bernie Sanders has separately proposed a one-off tax of 50 percent on the equity of major AI firms, framed as compensation for the American workers and infrastructure that made the industry possible. That is a very different mechanism — a coerced levy rather than a voluntary donation — and it comes with a very different political constituency. Sanders and Altman have reportedly discussed the sovereign wealth fund concept directly, which is part of why the FT described Altman’s proposal as an attempt to “buy goodwill” before something worse landed. Cybernews described it as “5 percent for U.S. government under Trump amid AI profit grab debate,” which captures the pressure Altman is trying to release valve.

The pressure is not just from the left. In June 2026, the White House asked OpenAI to limit the release of its upcoming GPT-5.6 model to a small number of government-approved partners, citing advanced capabilities. Anthropic’s Claude Fable 5 was pulled offline entirely for 19 days on export-control grounds before being restored on July 1 with added cybersecurity classifiers. The Five Eyes intelligence alliance issued a joint June 22 statement warning that frontier AI capabilities are “anticipated to exceed current industry expectations” within months, not years. That is regulatory oxygen the entire industry is now operating in.

Trump himself canceled a planned Oval Office signing ceremony for a new AI executive order in late June, telling reporters he did not want to do anything that would interfere with the U.S. competitive position in AI. That cancellation was reported to be tied specifically to concerns raised by Anthropic’s Mythos model in Five Eyes-linked cybersecurity assessments. AI labs are learning that Washington now has functioning leverage on release timing, export controls, and pricing. The equity discussion is downstream of that.

Altman’s pitch is not just about OpenAI. The FT reported he specifically named Google, Meta, Anthropic, and xAI as companies that would each contribute 5 percent of their equity to the fund. As of the July 2 report, none of those companies had publicly signaled support. Anthropic told Reuters directly that the U.S. government and Anthropic have not discussed the government taking a stake in the firm. That is a clean, direct denial from Dario Amodei’s team.

Anthropic has, however, been pushing a parallel policy: a “digital dividend” concept, laid out in a June paper calling for “universal pre-distributive capital accounts” funded by taxes on the AI sector, with initial priority given to workers displaced by AI-related job disruption. That is closer to Sanders’s approach than Altman’s — taxation, not donation — and it lets Anthropic support the goal of broader public benefit without volunteering the equity itself. Google, Meta, and xAI declined to comment when TIME sought a response.

The uncomfortable truth for OpenAI is that a voluntary 5 percent giveaway from four rivals is a very heavy ask, and Altman has effectively volunteered them. Each of those companies has its own venture and public-market investors who would have to accept the dilution. Meta is publicly traded and answerable to its shareholders. Google is inside Alphabet, which has its own capital allocation obligations. Anthropic just raised at a valuation north of $150 billion and has different political relationships. xAI has Elon Musk, whose personal conflict with Altman is well documented from the Musk v. Altman trial that concluded in May 2026. None of that says “voluntary donation.”