Anthropic Donations Top $540M as IPO Approaches

Anthropic is on track to become one of the largest corporate donors in the United States, and most people outside the company have not noticed. The money is not coming from a corporate foundation or a marketing budget. It comes from employees giving away company shares, with the company adding more shares on top.

The details come from documents Anthropic shared with potential investors ahead of its planned stock market listing, which The Information obtained. They show a donation program that is already very large, and one that could grow much larger once the shares trade publicly.

How the matching program works

The mechanism is simple. Anthropic employees can donate their shares to charities. The company then matches those gifts with additional shares of its own. The match is not the same for everyone: early employees receive three times the value of their donation in company shares.

The program has become expensive. Between October 2025 and March 2026 alone, the matching cost Anthropic more than $660 million. Because the company pays in newly allocated shares rather than cash, the expense falls on everyone else holding equity. Each matched donation dilutes the stakes of other investors, which matters to anyone considering buying in at the IPO.

The documents confirm a point that is easy to miss: what looks like an internal perk is also a recurring cost on the cap table.

The numbers against corporate America

Total donations in 2025 reached $540 million. To put that in context, the next-largest donors among Fortune 500 companies were Truist Financial at $115 million and BlackRock at $109 million. Anthropic's figure is close to five times either of them.

Those two firms are established financial institutions with long histories of corporate giving. Anthropic is a private AI developer that, according to leaked IPO filing details, is still running a large operating loss while growing quickly. A company that is not yet profitable is giving more than established banks and asset managers.

The figure could also rise sharply. If Anthropic's stock price keeps climbing after the listing, the value of donated and matched shares climbs with it. According to the report, total donations could then reach the billions.

Founders pledge most of their wealth

The giving is not limited to rank-and-file staff. Anthropic's IPO documents state that CEO Dario Amodei and the company's six co-founders have pledged to give away at least 80 percent of their wealth. For founders of a company preparing for one of the largest tech listings in recent memory, that commitment could translate into very large sums.

An Effective Altruism culture

Part of the explanation is cultural. Many Anthropic employees follow the principles of Effective Altruism, a philanthropic movement that tries to direct money to the causes where it can do the most measurable good. The movement has long had ties to the AI safety community, and Anthropic is often seen as close to that world.

According to former colleagues cited in the report, employees are already discussing in chat groups where the millions they expect to receive should go. The causes that come up most often are:

1. Global poverty - a long-standing focus of Effective Altruism, usually centered on interventions with measurable outcomes.
2. AI safety - work aimed at reducing risks from advanced AI systems, the field Anthropic itself was founded around.
3. Animal welfare - another area the movement has traditionally prioritized.

That this planning is happening before the IPO shows how much employees expect their shares to be worth.

What to Watch

For readers following the AI industry, this matters for two reasons.

The first concerns investors. A matching program that cost more than $660 million in six months is a meaningful source of dilution, and its cost moves with the share price. As AI labs line up for public markets, with OpenAI also preparing an IPO, it is worth watching whether Anthropic caps, changes or explains the program more fully once it faces public shareholders and quarterly scrutiny.

The second concerns influence. If donations reach the billions, a substantial new pool of philanthropic money will flow toward AI safety research and related causes, largely guided by people who build frontier models. This suggests that the safety field could become more dependent on money from inside the industry it studies. That is not necessarily a problem, but it raises questions about independence that deserve attention. The open question is how the giving will be distributed once the shares can actually be sold.