--- title: "Anthropic Public Benefit Fund: 5% of shares for public good" description: "Anthropic is committing 5% of its shares to a new fund for societal good. How the Public Benefit Fund works, where the money may go and what is still unknown." date: 2026-10-20 category: AI tags: ["AI", "Anthropic", "Claude", "labor market"] url: https://uper.pl/en/blog/anthropic-public-benefit-fund/ --- # Anthropic gives 5% of its shares to the public good — how the Public Benefit Fund works and what is still unknown **Anthropic, the company behind the Claude models, is committing 5% of its shares to a new fund for societal good, the Anthropic Public Benefit Fund.** The company announced it on October 18, 2026. Stockholders still have to approve the transfer, and the fund's stake can grow along with the company's valuation. The first area Anthropic points to is the effect AI is having on the job market. ## How the Anthropic Public Benefit Fund works Instead of an annual giving budget, the fund receives equity in the company itself. Anthropic describes it as a structure whose stake grows as the company grows. The mechanism has two parts: - **initial stake:** 5% of Anthropic's shares, transferred once stockholders approve, - **additional grants:** the board may give the fund another 2% of shares for every 50% increase in the company's value above a set baseline. The board decides on the additional shares, so the fund does not receive them automatically. Anthropic has not said where the baseline sits or how the later thresholds will be measured. ## Does the fund pay for AI safety research? Some coverage presents the fund as a source of money for AI safety and ethics research, but Anthropic's announcement says nothing of the kind. The company states plainly that most of its public benefit work stays in-house. It lists: - AI safety and **interpretability** research, which studies how a model arrives at its answers, - applied biomedical research, - discounted Claude access for nonprofits and scientific institutions. The Public Benefit Fund is meant to complement that work and allow "larger or different" approaches than the ones that fit inside the company's day-to-day operations. ## Where the money could go Anthropic has not published a list of programs, only examples of possible directions. ### Large donations to nonprofits The first direction is traditional philanthropy: large gifts to established nonprofit organizations. It is not known whether there will be an open call for applications or whether the fund will pick recipients itself, or what criteria it will use. ### Direct distributions to specific groups The second direction is money paid directly to particular groups of people, with no intermediary organization. The company has not said who could receive it, how much or on what terms. ### The impact of AI on jobs Anthropic names only one area outright: the effect of AI on employment. It notes that it has already committed $200 million to this area through its **Economic Futures Research Fund**, and adds that given the pace of AI progress "bolder approaches may be needed", which the new fund is meant to have the flexibility to pursue. The announcement does not promise any specific program, such as compensation or retraining for workers. For now the job market is a direction, and the plan has yet to be written. ## Why shares and not cash The most unusual part of the initiative is that the fund receives company equity rather than transfers from profits. Anthropic explains it with a comparison: tech companies usually reward executives with ever-larger equity packages, and Anthropic wants to give part of that equity to the public good instead. Two things follow from that: 1. **The fund's value depends on Anthropic's valuation.** When the company grows, so do its holdings, and past each threshold it may receive more shares. 2. **Shares cannot be spent like cash.** Anthropic is a private company, so to fund actual projects the fund will need some way of turning equity into money. The company has not explained how yet. The weak point of this model is that money for social causes depends on the commercial success of a single company. If the valuation falls, the fund will be able to do less. ## Why an LLC and not a 501(c)(3) The fund will be a limited liability company (LLC) wholly owned by Anthropic. The company deliberately chose not to make it a **501(c)(3)**, the US status for tax-exempt charities. Anthropic's reason is flexibility: an LLC can combine different kinds of giving, from donations to organizations to direct payments to people. In the US, 501(c)(3) organizations face stricter rules on who they can give money to and for what. The announcement says nothing about the tax consequences of the choice. The company did announce one clear restriction: equity held by the fund cannot be used for lobbying or political giving, in the US or abroad. How that ban will be enforced will only become clear once the governance rules are published. ## How much of Anthropic is going to the public good The new fund adds to two earlier commitments: - **the co-founders** pledged at the company's founding to donate 80% of their equity, - **employees** can donate their shares to charitable causes, and Anthropic matches those donations. According to the company, roughly half of its staff take part. Together with the fund, that puts roughly 20% of Anthropic's shares toward the public good, with more possible as valuation milestones are reached. Keep in mind that the 20% is a pledge. Most of that value will reach anyone only once the shares are sold or transferred. ## Context: a crisis of trust in the AI industry TechCrunch notes that the announcement comes as public sentiment toward the AI industry keeps getting worse. Anthropic CEO Dario Amodei has previously described the situation as "fundamentally a crisis of trust". Anthropic itself has had a rough few months, as I covered when [the US government ordered Claude Fable 5 pulled](/en/blog/fable-5-pulled-by-us-government/). That is why AI companies increasingly talk not only about model safety but also about sharing the gains from automation. Whether the fund helps Anthropic win back trust depends on how transparently it operates and what it actually pays for. ## What it means in practice - **Nonprofits and research institutions** may gain a new funding source, though for now there is nowhere to apply, because no procedures or criteria have been announced. - **Workers whose jobs are being reshaped by automation** have heard that the company takes the problem seriously. There is no concrete help yet. - **Other AI companies** get a reference point. It is now easier to ask them how much of their own equity goes toward softening the effects of the technology they sell. ## What we still don't know - **Governance:** who will decide on spending and who will oversee it. TechCrunch calls this the main open question. - **Valuation thresholds:** the baseline from which a 50% increase in value is counted. - **Turning shares into money:** how and when the fund will sell its holdings. - **Recipients and criteria:** who will get support and on what terms. - **Stockholder approval:** the initial 5% transfer is still awaiting it. In the coming months Anthropic is expected to publish details on the structure, governance and valuation milestones, along with a public benefit report on its progress toward its mission. ## Summary - The Anthropic Public Benefit Fund starts with 5% of Anthropic's shares, and the board may add 2% for every 50% increase in valuation above a baseline. - AI safety research stays inside the company. The fund is meant for work beyond its normal operations, above all responses to the impact of AI on jobs. - The LLC structure leaves room to choose how to give, and the ban on lobbying and political giving is so far the only firm restriction. - Together with earlier commitments, about 20% of the company's shares are set to go to the public good, though for now that is only a pledge. - The fund can only be judged once the governance rules and the first spending decisions are published. ## Sources 1. **Anthropic — Investing in the public benefit** [https://www.anthropic.com/news/public-benefit-fund](https://www.anthropic.com/news/public-benefit-fund) 2. **TechCrunch — Anthropic says it’s committing 5% of its shares to a ‘public benefit fund’** [https://techcrunch.com/2026/10/18/anthropic-says-its-committing-5-of-its-shares-to-a-public-benefit-fund/](https://techcrunch.com/2026/10/18/anthropic-says-its-committing-5-of-its-shares-to-a-public-benefit-fund/)