One of the more reassuring recent developments in Silicon Valley is that the people who might destroy humanity have begun setting aside money to compensate the survivors. Thank heavens!
Last Sunday, AI company Anthropic announced plans to allocate 5% of its shares to a philanthropic fund, potentially worth around $75bn. The timing is pertinent. Anthropic is preparing for a stock market listing that could value it at more than $1-trillion, and its IPO warnings contemplate “catastrophic or existential risks to humanity” from the very technology it sells.
A prospectus promising both fabulous wealth and possible extinction presents investors with a novel risk-return calculation. The consolation, apparently, is that if the machines take over, somebody might receive a charitable grant.
Elon Musk responded with a competing offer: next year he would give one SpaceX share to every American child. At about $146 apiece, the promise amounts to roughly $10.5bn — not that we are counting, but that’s less generous than his competitors. Musk has long been sceptical about conventional philanthropy, though he does have a substantial foundation. His announcement is less a conversion than an innovation: philanthropy by social media post.
Neither organisation is entirely new to generosity. Anthropic’s founders made charitable pledges years ago. But the scale of these commitments, alongside newly public SpaceX and Anthropic’s approaching IPO, recalls one of Mark Twain’s jokes.
In 1905, the American Board of Commissioners for Foreign Missions was debating whether to accept money from John D Rockefeller, whose business practices had made him almost as unpopular as he was rich. Twain wrote a letter to Harper’s Weekly in the voice of Satan, outraged that Rockefeller’s money should be regarded as tainted.
“The American Board accepts contributions from me every year: then why shouldn’t it from Mr Rockefeller?"
Satan’s point was that charitable donations often originate in questionable fortunes. But, you know, sometimes they don’t; sometimes they originate from great businesspeople who are hugely talented and who just want to make a contribution. Rockefeller was arguably simply unfortunate enough to have his sins discussed in the newspapers.
Peter Buffett, Warren Buffett’s son, revisited the problem in a 2013 essay called “The Charitable-Industrial Complex”. The wealthy, he argued, benefit from an economic system that produces inequality, then win applause for donating a fraction of their fortunes to relieve its symptoms. He called this “conscience laundering” — an idea that might have occurred to the founders of Anthropic. He also criticised “philanthropic colonialism”: donors descending on communities they barely understand, armed with chequebooks and an impressive certainty about what everyone else needs.
More to the point, what precisely are Anthropic and Musk proposing? Anthropic proposes putting equity into an institution that can finance charities or direct payments. The capital remains under institutional control. Musk would give shares directly to individuals, who could benefit from appreciation and decide what to do with their assets. One approach funds assistance; the other transfers ownership. Neither guarantees a useful result.
South Africa is ahead in the game on all this, as we know. And the results can be fabulous —and they can be disastrous. Just to take one example of many, in 2005 South Africa’s longest-listed black-owned empowerment company Brimstone created its Empowerment Share Trust, awarding shares to charitable organisations rather than recurring grants. By 2025 it had allocated more than 2.4-million shares to 42 nonprofits. The organisations received dividends and the prospect of capital growth. Brimstone’s lacklustre share-price performance has limited that second benefit, but the model showed charities could become asset owners instead of permanent supplicants. Many of the charitable institutions who received donations haven’t sold their shares.
Sasol supplies one cautionary counter-example among many. Its 2008 Inzalo empowerment scheme financed share ownership rather than outright gifts. When parts of the scheme wound up in 2018, financing costs had absorbed the value for employee and group participants. Employees had received dividends along the way, but no residual shares. It is possible, apparently, to give somebody a financial instrument that spends a decade aspiring to become a share.
But though direct ownership is no panacea, it does avoid lots of pitfalls. Charities can finance schools, clinics and other things individual shareholders cannot. But there is a fundamental difference between giving poor people money and giving them wealth. Money can be spent once. Wealth can, with a little luck, be spent by your grandchildren.
If the primary objective is reducing wealth inequality, direct individual ownership has the strongest conceptual justification. You are changing who owns the capital. If the objective is delivering essential services, charities may be better positioned to use large amounts of money effectively. A village cannot construct a functioning water system merely by giving its residents small shareholdings.
If one wants to be cynical, and I would never stoop so low, you could say corporate philanthropy has become an industry in which billionaires acquire the privilege of deciding how to remedy the inequality that made them billionaires. As Martin Luther King Jr put it: “Philanthropy is commendable, but it must not cause the philanthropist to overlook the circumstances of economic injustice which make philanthropy necessary.”
But I think it’s diversionary to get all caught up in the motivation. In the end, there is something encouraging about the possibility that the next generation of extraordinary fortunes might incorporate philanthropy into the structure of the businesses right from the start, rather than waiting for their founders to grow old, contemplative and concerned about their reputations. Anthropic’s decision to set aside a substantial equity stake could create an immensely valuable source of funding for public purposes, while Musk’s proposal introduces millions of children to the ownership of capital.
Neither is perfect, but there is a refreshing ambition about both.
