Sam Altman and the White House have been talking about a government stake in OpenAI for more than a year , with the AI startup potentially donating equity to seed a "Public Wealth Fund" that would give American citizens a slice of AI-driven growth. OpenAI is valued at more than $850 billion by private investors , and the Trump administration has already taken stakes in Intel, IBM, and other quantum and critical mineral companies . The concept sounds bold—Silicon Valley meets sovereign wealth—but the real story this week isn't who owns a piece of OpenAI. It's who's willing to keep paying for it.
A new spending discipline is taking hold inside corporate America, as CFOs and boards crack down on inefficient AI spending and question whether every task needs the top frontier models . The solution emerging is model routing, a tool that matches the job to the model, sending hard problems to expensive frontier models and easy ones to cheaper, faster alternatives . Glean CEO Arvind Jain estimates that roughly 95 percent of enterprise AI usage still runs on the most expensive frontier models, even for work cheaper models could handle . That inefficiency is about to end—and when it does, the trillion-dollar valuations underpinning the AI boom will face their first serious stress test.
Can OpenAI Justify Premium Pricing When Customers Route Around It?
Cisco's chief product officer laid out the math: at roughly $200 of token usage per employee per week, that's about $10,000 a year per person—meaning a 90,000-employee company is looking at $900 million annually . Uber burned through its entire 2026 AI budget in four months as Claude Code adoption jumped from 32% to 84% of its 5,000-engineer organization, with monthly API costs per engineer ranging from $500 to $2,000 , according to The Information.
The response isn't to stop using AI. Cognition CEO Scott Wu says companies can get about five to ten times better cost efficiency on routine work by using smaller models . Wu gave the example of asking a model to name the third U.S. president—each one, no matter how expensive, will tell you it was Thomas Jefferson . Why pay frontier-model rates for that?
If companies steer easy, high-volume work to cheaper open-source models, then OpenAI and Anthropic stop getting paid for every task—they only get the more complex jobs, undercutting the premium-demand assumptions behind their valuations , CNBC reported. At standard rates, OpenAI is cheaper across most comparable tiers in 2026 and uniquely offers ultra-budget Nano models with no Anthropic equivalent , but Google's Gemini 3.5 Flash could save enterprises more than $1 billion a year in AI costs, with Sundar Pichai arguing that shifting 80% of workloads to a combination of Gemini Flash and frontier models would save over $1 billion annually , VentureBeat reported.
Why Is the White House Talking Equity Now?
Altman first pitched the idea to President Trump in early 2025 and has raised it again with senior officials in recent weeks , according to CNBC and NOTUS. One option would have OpenAI donate equity to seed a Public Wealth Fund that would give every citizen a stake in AI-driven economic growth . Trump addressed the possibility Friday, saying "there are concepts where pieces could be given to the American public, where the American public essentially becomes a partner" , according to the International Business Times.
Sen. Bernie Sanders called this week for the U.S. government to acquire 50 percent equity stakes in AI companies, effectively bringing them under federal control . Critics on both the left and right see downsides, with Public Knowledge warning the government would be "a shareholder and a regulator at the same time, which creates substantial conflicts of interest" , NOTUS reported.
The timing matters. Anthropic confidentially submitted draft IPO paperwork Monday, potentially leapfrogging OpenAI in the race toward a Wall Street debut as soon as this fall, after raising $65 billion at a $965 billion valuation that eclipsed OpenAI's value for the first time , Fortune reported. Anthropic's revenue run rate has ballooned to $47 billion, up from $10 billion in annual revenue last year , CNBC reported. Anthropic's confidential filing appears to be a strategic move to secure first-mover advantage and position itself as the industry's pricing benchmark, with consecutive outperformance in revenue, valuation, and profitability giving it the confidence to file first , according to analysis from Funn.
If OpenAI goes public with a government shareholder on the cap table, it changes the optics—and possibly the economics. A government stake would come with immediate governance questions, as federal officials who owned part of an AI company could be pulled between promoting competition, policing safety, and protecting taxpayer returns, blurring the line between regulator, customer, and investor , Prism News reported.
