Anthropic’s latest announcement over a “public benefit fund” has been met with plenty of skepticism online.
Anthropic has announced a new vehicle called the Anthropic Public Benefit Fund, a wholly owned entity that the company says will be dedicated entirely to “the public benefit.” The fund will start off holding 5% of Anthropic’s shares, and the company has said that figure can rise by up to an extra 2% for every 50% increase in the company’s valuation over a set baseline, subject to the board’s discretion. Combined with the fact that Anthropic’s co-founders have already pledged to give away 80% of their personal equity, and that roughly half of employees take part in a stock-donation matching programme, the company says around 20% of its total equity is now committed to “public” purposes.
The announcement lands just weeks before what is expected to be one of the largest IPOs in history, with Anthropic’s valuation having already crossed $960 billion earlier this year. The fund, structured as an LLC, is intended to make donations to nonprofits, fund “direct distributions to specific populations,” and address AI’s disruption of the labour market. Anthropic has said the money will not be used for lobbying or political donations, in the US or abroad, and that it will publish a public-benefit report detailing how the fund is used.
The reaction online, however, has been sharply critical — not of the idea of a public benefit fund itself, but of how little is actually known about who will control it, who will receive money from it, and what “public good” will end up meaning in practice.
What The Critics Are Saying
Elon Musk, who runs rival AI lab xAI, was among the first to respond, posting under Anthropic’s own announcement that he “recommend[s] giving the shares directly to the public” rather than routing them through a separate entity.
David Sacks, the venture capitalist and former White House AI czar, reacted to a thread from Anthropic co-founder Jack Clark noting that roughly 20% of Anthropic equity is now committed to “the public good.” Sacks’ response was blunt: Anthropic has “put 20% of its equity in an LLC,” and “whether it serves the public good remains to be seen” — a pointed way of saying the fund’s structure, on its own, proves nothing about its actual impact.
Palmer Luckey, the Oculus founder and Anduril CEO, framed his criticism around who the fund is actually meant to help. He argued that Anthropic’s idea of “public good” means helping “everyone else” rather than specifically the American public, and suggested the initiative “reeks of USAID 2.0” — a way, in his telling, to funnel money out of the US economy rather than reinvest it domestically.
Travis Kalanick, the Uber co-founder, was more sardonic, describing the announcement as “the EA infinite money glitch” — a reference to the effective altruism movement that has long been associated with Anthropic’s founders and early investors, and how it will likely keep propagating with this new fund.
Jason Calacanis, the All-In podcast co-host, took the most tongue-in-cheek approach, proposing — apparently as satire — that Anthropic instead launch a universal basic income programme paying Americans $3 a month, joking it would make the company “the most loved company in America” for a fraction of its profits.
X account PoliMath offered one of the sharpest readings of the fund’s likely purpose, calling it a potential “NGO slush fund” that could function as a way to win over left-leaning politicians and ensure they follow Anthropic’s lead on AI policy and regulation.
A Pattern Of Scrutiny
This isn’t the first time Anthropic’s structure and its ties to the effective altruism movement have drawn criticism from people across the political spectrum. The company’s close links to EA circles — co-founder Daniela Amodei is married to Holden Karnofsky, a co-founder of the EA-aligned Open Philanthropy and GiveWell — have long made it a target for accusations of running a closed, values-driven network rather than a transparent corporate structure. Sacks has previously accused Anthropic of running a “sophisticated regulatory capture strategy based on fear-mongering”.
Conclusion
What ties these reactions together is not disagreement over whether AI companies should give something back, but discomfort with the fact that nobody outside Anthropic actually knows how the Public Benefit Fund will decide who gets paid. The company has said the fund won’t go toward lobbying or political donations, but it has not said who sits on its board, how grant recipients will be chosen, or what oversight — if any — will exist beyond Anthropic’s own say-so. A pledge to “address AI’s impact on the labour market” or make “direct distributions to specific populations” can mean almost anything, and right now only Anthropic gets to define it.
That opacity is what should raise eyebrows. A company that is simultaneously the steward of one of the most powerful AI models in the world, a major political actor in Washington, and now the sole gatekeeper of a multi-billion-dollar philanthropic vehicle is concentrating an unusual amount of discretionary power in one place — power to decide which causes, populations, and organisations are deemed worthy of its backing, with essentially no external check on that judgment. Handing out billions of dollars’ worth of equity is not inherently dangerous; doing so through a structure that answers only to the company’s own board, with no disclosed criteria for who benefits, is. Until Anthropic says more about who actually controls the Fund and how recipients are chosen, it’s fair to treat this less as philanthropy and more as a company quietly deciding, on its own terms, who gets to share in its fortune — and who doesn’t.