OpenAI pitches 5% US government stake worth about $42.6B
TL;DR
- CNBC's headline frames the offer as a mechanism to 'ease Washington pressure,' diverging sharply from Altman's public-benefit framing of the same proposal.
- The structure is modeled on Alaska's Permanent Fund, casting the US government as a sovereign dividend recipient rather than a regulatory body.
- Altman pitched Trump, Lutnick, and Bessent directly, routing the proposal through executive decision-makers rather than regulatory agencies.
A US president being offered a 5% slice of the frontier AI lab that is arguably the most politically sensitive private company in the country is the kind of thing you would not have taken seriously a year ago. According to the Financial Times, Sam Altman has pitched exactly that: OpenAI would hand the US government a 5% equity stake, which at the company's recent $852 billion post-money valuation would be worth roughly $42.6 billion.
The pitch, as reported, is not a one-off conversation. Altman has discussed the arrangement with President Trump, Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent, and also with Democratic Senator Bernie Sanders. The broader framing is that Washington would take 5% of each of the leading US AI developers, with Anthropic, Google and Meta named in the reporting, via a sovereign wealth fund vehicle. That seeds what OpenAI has been calling a public wealth fund since an April policy paper, with returns from the AI build-out shared with citizens rather than concentrated in a handful of private cap tables.
If any version of this actually goes through, it rewires the politics of every future AI fight. Export controls, antitrust, model-safety rules, the coming battles over data centre siting and compute all look different when the Treasury has a direct financial interest in the incumbent labs winning. It also invites a Democratic counter-offer: CNBC frames the move as a way to defuse political blowback in Washington, and Sanders has separately called for a much larger 50 percent equity stake in leading AI companies, funded through a one-time stock tax, that would come with voting shares and board representation.
The honest caveats are stacked high. Anthropic has publicly confirmed it is not involved in these discussions. The reporting does not say what governance rights, if any, come with the 5%, and it does not say how the shares would actually be held given OpenAI's nonprofit-parent structure. Take the specifics as reported, not settled.
If the deal survives Congress, the labs that get in early buy political cover and the government gets an asset it did not have to pay cash for. Whether that is a good trade for smaller US AI startups, non-participating rivals and foreign labs is the question the reporting does not yet answer.
What others are reporting
-
CNBC Read →
Headline and URL slug ('address-political-blowback') frame the offer as political pressure relief rather than public benefit — the starkest editorial divergence from the FT's version.
-
Bloomberg Read →
Tier-1 financial press confirmation; 'FT Says' attribution in the headline signals Bloomberg had not corroborated the story through its own sources at publication.
-
Reuters Read →
Reuters explicitly noted it could not verify the report and that OpenAI and the White House did not respond — the clearest public signal the story was unconfirmed at wire time.
OpenAI has discussed giving a 5% stake to the U.S. government, the Financial Times reported on Thursday.
-
CNN Read →
CNN's international business desk pick-up gives the story global reach; 'intl' URL slug confirms it was filed for international audiences, broadening the story's political resonance beyond Washington.
-
Forbes Read →
Adds that Altman engaged Trump, Lutnick, and Bessent directly, and that the structure is explicitly modeled on Alaska's Permanent Fund, framing the government as a shareholder rather than a regulator.
OpenAI is in talks to potentially give the Trump administration a 5% stake in the company.
-
CoinDesk Read →
Frames the proposal as a potential industry-wide governance precedent and foregrounds that Anthropic, Google, and Meta buy-in is required but not secured, elevating the structural uncertainty.
The proposal, which remains in the conceptual stage, was reportedly raised by OpenAI CEO Sam Altman during early discussions with U.S. officials.
Shared on Bluesky by 4 AI experts
Originally reported by Financial Times
Read the original article →Original headline: OpenAI Proposes Giving US Government a 5% Stake — Would Be Worth ~$42.6B at $852B Valuation, Altman Reportedly Pitched Trump, Lutnick, and Bessent