Pop quiz. One artificial-intelligence giant announced a new vehicle for channeling roughly 5% of its equity to nonprofits and to the people whose livelihoods artificial intelligence is about to take. Another billionaire says he wants to give every American child one share of his rocket company by next year. Which is the real act of generosity?
Anthropic. Not even close.
Strip the showmanship away from the SpaceX giveaway and what is left is a voting bloc the size of Spain. While details of Mr. Musk’s plan remain vague, the idea appears to be to hand every child a single share in a private company whose valuation moves with every Musk tweet. The donors — that’s us, the public — get a tax deduction, or don’t, depending on fine print that has not been written yet. Mr. Musk gets the ballots.
The parents of roughly 73 million American children will hold, in aggregate, about 73 million shares of SpaceX stock at a price near $147 apiece — a stake worth roughly $10 billion, all of it voting stock. The children, until their eighteenth birthdays, are carriers for a position in a company that builds launch vehicles, builds satellites, competes for defense contracts, and sits at the heart of the most ambitious private space program in history. Those ballots will be cast by parents, tutors, and custodians on whom the gift lands hardest, in low- and middle-income households whose children now hold one share of a company the parent did not choose to invest in. Putting 73 million minors on a corporate roll is not philanthropy; it is fan mail with a stock ticker. A breathtaking transfer of corporate governance to shareholders with no say, no proxy, no ability to sell, and no recourse if the company falters. The math of who owns America’s political future just moved.
The proposal is unrelated to, but consistent with, the concept behind the new so-called Trump accounts, which the Treasury Department will seed with $1,000 for every child at birth. Michael and Susan Dell have committed $6.25 billion to fund $250 contributions to some 25 million Trump accounts — money the recipient may do with as he pleases, subject to a holding period that locks it into a politically curated vehicle, a structure with its own democratic and distributional problems that deserve separate scrutiny. The Dells, that is, are also locking the votes of minors into a politically loaded vehicle, exactly like Musk. The Musk plan is the same instrument in a louder wrapper, with a single donor at the center rather than many.
This is not philanthropy. Philanthropy ends with the recipient. This begins with the recipient and recurses back to the donor, in the form of voting control exercised over a company whose competitors, contractors, and regulators are themselves subjects of public policy. The donor’s allies prefer to call the gesture a magnificent charitable troll, a dig at the philanthropic establishment. What the device actually is, is a 73-million-share voting trust whose beneficiaries cannot consent and whose trustees have an obvious interest. Read the Musk pattern: the donations to political committees, the public scolding of lawmakers, the threats to move operations to jurisdictions more compliant with his preferences, and now the systematic distribution of voting stock to every child in the country, the explicit purpose being to make them stakeholders in his enterprise. This isn’t trolling. It is a long-position acquisition of political time.
Anthropic’s move is harder to fundraise against, precisely because it does what a real give-back is supposed to do. Roughly 5% of the company goes into a flexible limited liability company that may fund existing nonprofits or may explore direct distributions to specific populations — including, plausibly, the workers displaced by the very systems Anthropic is building. “Might explore” sounds modest only to readers who mistake modesty for unseriousness. An institution built to weigh need, scale, and timing is the opposite of the populist check-writing that dominates today’s mega-donor circuit. A fund whose design is responsive to damage as it appears is the structure a serious actor builds when its product will displace people. A fund whose design is locked to the donor’s private balance sheet, and whose recipients are politically captive, is what the alternative looks like.
Left unanswered are questions such as which existing nonprofits, to do what, and which specific populations. Those are questions of design, not disqualification. Serious institutions are built to answer governance questions rather than hide from them. Anthropic’s fund will be governed; Musk’s giveaway will not. The contrast is the story.
The defenses of the Musk plan collapse on inspection. We are told the device hands recipients ownership and choice. But children do not exercise voting rights at five, or ten, or fifteen. The votes sit in custodial accounts, voted by parents on whom the gift lands hardest. That isn’t capitalism with choice. It is private welfare administered by a single actor’s foundation, with no diversification, no oversight, and an explicit alignment of voting interest to a single political faction. And it is harder to defend than it sounds: ample Silicon Valley experience counsels skepticism about one-man philanthropy by another name, and giving away shares of a single volatile company to millions of children who cannot sell them is a stunt that ends up funneling political goodwill toward one celebrity donor whose real-world benefits are hard to discern.
Then comes the cynicism. The mantra that foundations funnel cash into politically well-connected causes whose real-world benefits are hard to discern has become conventional wisdom. Pick a politically unsavory grant by the Gates Foundation if you like, or point to the school-choice scholarships bankrolled by John Walton and Jeff Yass, which can change individual life prospects only when the surrounding infrastructure is sound — which it usually is not. But the careful disbursements of the Ford Foundation and the Open Society foundations have built genuine capacity in education, public health, and civic life, and by the same standard the Walton and Yass programs are measurable, evaluable, and answerable. Stated honestly, the argument is that private money in public life is suspect across the board. That is a critique that belongs on the Musk giveaway with ten times the force it gets aimed at the relatively accountable giving of the Gates Foundation. The Walton and Yass programs face measurable outcomes; Musk’s does not. The giveaway is a one-way transaction, not a philanthropic instrument with evaluable results.
Anthropic’s structure does have a flaw: it concentrates decision-making in a small group of executives who will have to defend, in public, every dollar of disbursement — who they pick, how much they pay, which nonprofits they fund. That is precisely why it is the more honest instrument. The scrutiny will be unbearable at times, and the political flack will be relentless. That is the test of seriousness. The donor whose gifts bypass the test is the donor who, history suggests, was never giving in the first place. The real charitable troll here is the one who treats every policy problem like a marketing opportunity.
