Anthropic pledges 5% of its shares to a public benefit fund, if shareholders say yes

The announcement published on Anthropic's website on 18 October is titled "Investing in the public benefit" and contains a figure and a condition. The figure is 5% of the company's shares, pledged to a fund called the Anthropic Public Benefit Fund. The condition comes down to three words, "Subject to stockholder approval": the pledge takes effect once shareholders approve it, and none of the sources I consulted reports that the vote has already taken place. Then there is a growth mechanism: the board may, at its discretion, add a further 2% of shares for every 50% increase in the company's value above a starting threshold that Anthropic has not disclosed. The logic is in the company's own words: "the more Anthropic grows, the larger the share of the company that can be dedicated expressly to the public good". TechCrunch (Anthony Ha, 18 October 2026) independently confirms the main terms.

The vehicle is an LLC wholly controlled by Anthropic, not a 501(c)(3) charity. The company says it chose this form for more flexibility across different kinds of giving: large donations to traditional non-profits, or direct distributions to specific groups of people, citing the effects of AI on the labour market as an example. One limit is spelled out explicitly: "None of the Anthropic equity held by the LLC will be used for lobbying or political giving, in the US or abroad." Anthropic is a Delaware public benefit corporation and already has a Long-Term Benefit Trust with a role in corporate governance; the announcement gives it no role in the new fund. The $200 million the company had already committed to studying and managing AI's impact on work is a separate sum.

On the numbers, it pays to be precise about what is stated and what is arithmetic. The announcement gives no dollar value. The only official reference point is the Series H of 28 May 2026: $65 billion raised at a post-money valuation of $965 billion. 5% of that comes to roughly $48.25 billion, but that is a multiplication, not a valuation of the fund; the estimates circulating around $100 billion rest on a hypothetical IPO that Anthropic has not confirmed. The company also estimates that around 20% of its shares will go to the public good, and that figure adds up three separate commitments: the fund's 5%, the co-founders' pledge to donate 80% of their stakes, and the matching programme for employee share donations, in which about half of the staff take part, plus any shares tied to valuation milestones. The essentials remain open: who will run the fund, whether there will be independent oversight, what criteria will select the beneficiaries, what threshold triggers the 2%, and when the first payouts will arrive. Anthropic says details on structure, governance and thresholds will follow "in the coming months". Forbes reports criticism of the LLC choice: I could not check it at the direct source because the page was not accessible, and I flag that as a gap in my piece.

What interests me here is not the percentage, which is high, but the distance between the announcement and the fact. The announcement contrasts the fund with the tech-industry habit of handing executives ever-larger stock awards, and the line quoted by TechCrunch — "given the pace of AI progress, bolder approaches may be needed" — explains the why of the initiative; the how is still to be written. A pledge that hinges on a vote whose outcome is not yet on record, with an unpublished threshold and an unnamed manager, is a well-written proposal. The 5% becomes a fact on the day shareholders vote and the first payout arrives; until then, the most useful document is not the announcement but the one due in the coming months that will say who decides.

— Olya