A Moment for Pause

This week, SpaceX priced its IPO at $135 per share — a $1.77 trillion valuation, the largest public offering in history, raising $75 billion in a single day. Hot on its heels, Anthropic raised $65 billion at a $965 billion valuation and is targeting a public listing before year's end. OpenAI is eyeing a $1 trillion IPO that could dwarf even that. Together, these three companies are valued at approaching $4 trillion and are expected to raise a combined $200 billion from public markets.

“The question of who they’re accountable to — and who shares in what they build — is one we can no longer afford to treat as secondary.”

What these companies have actually built is genuinely staggering. SpaceX has done what most aerospace engineers once thought impossible — repeatedly landing orbital rockets, building global satellite internet from scratch, and seriously advancing humanity's capacity to become a multi-planetary species. Anthropic and OpenAI have compressed decades of AI research into just a few years, producing tools that are already reshaping medicine, scientific discovery, education, and the nature of knowledge work itself.

These are not incremental achievements. They are civilizational.

And yet.

Elon Musk's net worth crossed $970 billion today — largely on the back of his ~42% SpaceX stake — making him the world's first trillionaire. A single early investor holds a ~$68 billion position in SpaceX alone. Anthropic's seven founders each saw their personal stakes hit roughly $8 billion in a single day — the largest single-day addition of billionaires from one company in Bloomberg Billionaires Index history. The combined net worth of founders and major shareholders of just five AI-adjacent companies grew by approximately $1.2 trillion between January 2024 and early 2026 — a wealth accumulation rate exceeding any previous period in history, including the original Gilded Age.

None of this happened in a vacuum. These companies were built on publicly funded university research, decades of government contracts, tax-subsidized infrastructure, and the intellectual labor of tens of thousands of people who will see a fraction of this wealth.

Which brings me to a harder question.

Should wealth of this magnitude — generated at this speed, from technology this foundational — be governed primarily by shareholder value? Or have these companies crossed a threshold where they're better understood as public goods?

““Reward” and “unchecked accumulation” are different things. At some scale, private control of civilizational infrastructure stops being a success story and starts being a governance problem.”

It's worth noting that both OpenAI and Anthropic incorporated as Public Benefit Corporations, a structure that legally requires directors to balance profit with broader social benefit. OpenAI's stated public purpose is literally "to ensure that artificial general intelligence benefits all of humanity." Anthropic's founders have pledged to donate 80% of their personal wealth. These are meaningful commitments — and they deserve credit for making them. But a pledge is not a policy. And a PBC filing doesn't automatically determine how value flows when a company is worth a trillion dollars.

AI models are fast becoming what roads, electrical grids, and telecommunications networks once were — essential infrastructure that underpins nearly every economic and civic activity. Space-based internet is already providing global connectivity that governments couldn't build. When infrastructure reaches this level of ubiquity and dependency, economists and legal scholars increasingly argue that the standard shareholder-primacy model breaks down. Not because the founders don't deserve reward — they do — but because the implicit social contract that makes these companies possible demands something in return.

The questions worth sitting with:

• If AI becomes as essential as electricity, should access to it be treated as a utility — regulated, universally available, and priced accordingly?

• If these companies were built on publicly funded research and government subsidies, what does society have a legitimate claim to in return?

• Is there a point at which a single individual holding $970 billion — while the infrastructure that created it relied on public institutions — represents a structural failure of how we distribute the gains of innovation?

I don't think the answer is to punish ambition or penalize risk-taking. The founders of these companies took extraordinary bets on ideas most institutions wouldn't touch. That deserves real reward.

But "reward" and "unchecked accumulation" are different things. At some scale, private control of civilizational infrastructure stops being a success story and starts being a governance problem.

We are at that scale now.

These are remarkable institutions doing genuinely important work.

The question of who they're accountable to — and who shares in what they build — is one we can no longer afford to treat as secondary.