Silicon Valley billionaires' new approach to philanthropy: from donating cash to donating equity

More than a decade ago, a young, handsome man who had graduated from Oxford University stood in front of a grocery store shelf, hesitating over which breakfast cereal to buy. He was calculating in his mind whether, if he stuck to buying the cheaper brand for a year, the money he saved would be enough to save a person's life.

This man was the British philosopher William MacAskill. He later became an important advocate of Effective Altruism. The core of this set of ideas is to invest limited resources in things that can produce the greatest social benefit, and to use rational analysis and effect evaluation as much as possible so that every sum of money helps more people.

Today, Effective Altruism has become an influential current of thought in Silicon Valley's tech circles. Especially for some entrepreneurs and investors in the AI field, they believe that technological progress can not only create enormous wealth but may also bring major risks concerning humanity's future. How to ensure AI safety and how to respond to the impact of technology on employment and social structure have begun to become issues of concern to this group of tech elites.

America's new generation of tech tycoons has already taken some measures. Anthropic, the AI company about to become the next largest IPO in history, announced a special arrangement one week before listing: it will put 5% of the company's shares into the newly established Anthropic Public Benefit Fund. Based on outside expectations for its IPO valuation, the potential value of this equity could reach tens of billions of dollars, but this is not a cash donation that has already been realized. The fund may also receive more shares as the company's valuation grows, at the discretion of the board of directors. Anthropic said the specific valuation benchmark and governance arrangements will be disclosed later.

The next day after the news, the world's richest man, Musk, said on X that he planned to give every child in the United States one share of SpaceX stock starting in 2027, and called it only "the first of many planned steps." Based on the scale of about 72 million minors in the United States and the price of about $146 per SpaceX share at the time, the plan has a potential value of about $10.5 billion.

These tycoons who have created enormous wealth through technology are now loudly caring about humanity's fate, and they attempt to use charitable behavior to provide legitimacy for their wealth capture. These practices differ from traditional charitable behavior. Rather than simply returning wealth to society directly as a resource, as tycoons in the past were keen to do by donating money, donating supplies, building schools, and building libraries, they often choose social problems with long-term potential returns, set measurable goals, amplify influence through capital, technology, and talent, allow failure, and adjust resource allocation based on results.

Does it sound a bit like doing venture capital?

And this is precisely the core of Effective Altruism: resources should flow to places that can produce the greatest social benefit. But the problem is that the benefits and risks of technological progress are not always borne by the same group of people. An AI technology may significantly increase overall social productivity, and it may also cause some workers to lose their jobs. Simply calculating the total benefits brought by technology cannot answer who should bear the losses, or how the fruits of growth should be shared.

This also leads the question to a deeper level: when a small number of companies control AI models, computing power, and capital, can ordinary people only wait for companies or tycoons to decide how to compensate them? Can they participate in the distribution of wealth created by technology, or even become the owners of that wealth?

When donations become part of corporate capital structure

On October 26, Anthropic officially publicly submitted a prospectus to the U.S. Securities and Exchange Commission (SEC), planning to list on Nasdaq under the ticker symbol ANT. According to estimates by market institutions and investors, its IPO valuation may reach $1.5 trillion to $2 trillion. If this valuation is realized, Anthropic is expected to set a global capital market IPO record, and the company would even exceed the annual GDP of many countries in the world.

Anthropic mentioned in its prospectus that the AI technology it is developing may pose catastrophic or existential risks to humanity. The document shows that the company's co-founders have committed to using 80% of their personally held Anthropic equity for charitable causes. In January this year, founder Amodei said in an open letter that rich people have a responsibility to help workers and their families who will be left behind in the AI era.

As mentioned earlier, Anthropic proposed establishing a public benefit fund one week before the prospectus was made public, but the fund is established as a limited liability company rather than a traditional nonprofit organization, in order to retain flexibility among different donation methods. The company also said that the equity held by the fund will not be used for lobbying or political donations.

Musk, who also created one of the largest IPOs in history, does not agree with Anthropic's arrangement. After Anthropic announced its public benefit fund plan, he directly replied under the relevant post: "I suggest just giving the shares directly to the public." So Musk proposed his stock giveaway plan, but this is a personal commitment, and specific implementation details have not yet been announced, including the eligibility scope of the children receiving the shares, the method of stock transfer, and whether a lock-up period will be set.

Unlike Anthropic, which manages equity uniformly through a public benefit fund, Musk proposed directly distributing corporate shares to the public. What recipients get is not just one-time cash; they will also become shareholders of SpaceX, sharing in the capital gains the company may generate in the future while bearing the risk of stock price fluctuations.

Musk's promise also has a special background. In June 2026, SpaceX completed its listing, becoming one of the most closely watched IPOs in global capital markets that year, raising approximately $85.7 billion. After the company went public, the shares Musk holds gained a clearer market valuation. What he proposed to give away this time were SpaceX shares he personally holds, not funds provided directly by SpaceX as a company.

Before Musk, SpaceX President and Chief Operating Officer Gwynne Shotwell had also made a similar arrangement. In July 2026, she and her husband pledged to give SpaceX shares to more than 2 million children, mainly targeting low-income families in Texas. Based on the share price at the time, the donation was valued at about $320 million. Reports show that these shares accounted for about one-fifth of the Class A shares Shotwell holds or has the right to purchase. Her donation is far smaller than Musk's, but she likewise chose to transfer corporate equity directly rather than donate cash.

Compared with Anthropic and SpaceX, Dell Technologies founder Michael Dell's approach is much more straightforward. Last year, he and his wife Susan Dell pledged to donate $6.25 billion to inject $250 each into investment accounts for 25 million children in the United States. These accounts belong to the Trump Accounts program launched by the U.S. government, aimed at establishing long-term investment accounts for children. The Dell couple's funds are mainly directed to children in eligible low-income areas.

Unlike Musk's gift of SpaceX shares, what the Dell couple pledged is cash funding, which enters children's accounts through a family philanthropic arrangement, not direct allocation by Dell Technologies. The account funds will be used for long-term investment, and the children can obtain the corresponding assets once they reach adulthood. This approach combines charitable donation with personal asset accumulation; beneficiaries receive not only immediate assistance but also the opportunity to share in the long-term returns of capital markets.

Looking at these donations together, we can see that some new changes have emerged in the philanthropic methods of Silicon Valley tech billionaires.

First, donated assets have expanded from cash to corporate equity. The Dell couple put up cash, Musk and Shotwell chose to transfer SpaceX shares, and Anthropic placed company shares into a public benefit fund. The value of equity changes with corporate development, so the scale of donations is also tied to the company's future growth.

Second, the timing of donations has moved earlier. Traditional philanthropy often occurs after wealth has been accumulated, with billionaires transferring already-realized wealth to foundations or nonprofit organizations. Anthropic's approach, by contrast, puts a portion of equity into a philanthropic arrangement while the company is still growing rapidly and has not yet completed an IPO. Donation is no longer just the distribution of wealth at the endpoint; it is also beginning to become part of a company's capital structure.

Furthermore, the way benefits are delivered is no longer limited to traditional charitable institutions. The Dell couple inject funds into children's investment accounts, Musk proposed giving stocks directly to children, and Anthropic leaves room for future distribution of funds to specific groups. Beneficiaries have the opportunity to directly hold assets and participate in the long-term growth of enterprises and capital markets.

This change is not just the personal choice of a few billionaires. OpenAI's corporate structure provides another reference. After completing its restructuring in 2025, the original nonprofit organization was renamed the OpenAI Foundation and holds about 26% of the equity of the for-profit OpenAI Group. Based on the valuation at the time, this equity was worth about $130 billion. The foundation also retained control over OpenAI Group, with the power to appoint and replace board members. This is not a founder recently donating a sum of wealth, but rather a corporate governance structure that allows a nonprofit organization to hold equity in a commercial company and participate in decisions about the company's direction.

Who decides how ordinary people's wealth is distributed

Billionaires' enthusiasm for philanthropy is nothing new. In 1889, steel magnate Andrew Carnegie published "The Gospel of Wealth," putting forward the later widely known view: "He who dies thus rich dies disgraced."

In Carnegie's view, industrial society inevitably brings concentration of wealth. A few people accumulating enormous wealth through business ability is not inherently a bad thing; the real question is how that wealth is ultimately disposed of. He opposed rich people leaving their fortunes to descendants to squander, and also disapproved of waiting until death to donate wealth. He argued that the rich should, during their lifetimes, take on the responsibility of wealth stewards, using private wealth for public undertakings such as libraries, universities, and hospitals, turning wealth into social resources.

This set of ideas laid an important foundation for modern American philanthropy. Thereafter, industrial giants such as Rockefeller and Ford established large foundations one after another, and private wealth began to enter the public sphere in an institutionalized way. A logic that continues to this day was thus formed: wealth can be created and possessed by a few, but the rich have a responsibility to decide how to give part of it back to society.

More than a century later, this logic is changing.

The philanthropy of traditional industrial giants mainly centered on social giving back after wealth accumulation. They improved public services by donating to build schools, hospitals, and public facilities. But today's tech billionaires face a different situation. The technologies they create may themselves become forces that change wealth distribution and social structure.

Especially AI. It is not just a new commercial track; it may also change enterprises' production methods, workers' employment opportunities and sources of income, and readjust the relationship between capital and labor. This has also given tech billionaires new ideas about philanthropy: they believe they not only have the right to decide how money is made, but should also have the right to decide how money is spent and what problems society should prioritize solving.

This way of thinking is inseparable from a long-standing technological optimism in Silicon Valley—a belief that technological innovation can solve social problems, and that a small number of people with specialized knowledge, capital, and execution capacity can drive social progress more effectively than traditional institutions. In such a value system, entrepreneurs are not only creators of wealth but are also assigned the role of solvers of social problems.

Effective Altruism further provides a methodology for this role. It emphasizes that doing good cannot be based solely on the donor's subjective intentions; one must also measure how much actual effect the funds can produce. Different charitable projects should be compared on cost and benefit, and resources should be directed to areas that can generate greater social benefit. This set of ideas introduces scientific research, data analysis, and investment-return-style evaluation methods into charitable decision-making.

Longtermism extends this thinking to a more distant future. It holds that the interests of future generations are equally worthy of attention, and that humanity's present choices may affect its fate for hundreds of years or even longer. William MacAskill systematically expounded this view in What We Owe the Future. According to this logic, issues such as AI going out of control, major technological risks, and humanity's long-term survival may have higher priority than many immediate social problems.

Combined, these two sets of ideas have changed the way charitable resources are allocated. Donors no longer simply give money to traditional charitable organizations and then let the recipients decide how to use it; instead, they try to intervene in the entire process like investors: identifying problems worth solving, judging technological paths, screening execution teams, measuring project effects, and continuing to invest based on actual progress.

This is also the core feature of so-called "philanthrocapitalism." It brings the business world's management methods, performance standards, and investment logic into the charitable field. Charity is no longer merely a transfer of wealth from private accounts to the public sphere; it has also begun to become a way of actively allocating resources and influencing the direction of social development.

Therefore, the charitable actions of tech billionaires do not take place entirely outside business. They are closely connected to their sources of wealth, industrial interests, and judgments about the future. Anthropic's contribution of part of its equity to a public benefit fund is a response to AI's social impact and also a way of arranging in advance the distribution of future returns within the company's capital structure. Musk's proposal to give SpaceX shares to children similarly attempts to link personal wealth with public sharing in corporate growth.

But when billionaires begin to actively participate in the distribution of social wealth, who exactly gave them such power?

Rob Reich, a professor of political science at Stanford University, argues in Just Giving: Why Philanthropy Is Failing Democracy and How It Can Do Better that large private charitable foundations do not inherently conform to democratic values. They can convert private wealth into public influence without having to undergo elections, public deliberation, and the same degree of accountability as governments. Donors not only decide how much money to give but can also decide which problems are worth prioritizing, which research should receive funding, and which groups should benefit first.

This is precisely the most alarming aspect of tech philanthropy. The ability to create wealth does not naturally equal the qualification to decide the public interest. A person can become a billionaire through business judgment, but does not necessarily thereby have the legitimacy to judge the direction of society's development as a whole. Especially when AI may affect the work, income, and lives of hundreds of millions of ordinary people, the participation of technology owners in distribution through private foundations and public benefit projects cannot replace broader public discussion and institutional arrangements.

Ordinary people's participation in wealth distribution usually goes through institutional arrangements such as taxation, fiscal budgets, public policy, and social security. Although these mechanisms are imperfect, they at least involve public discussion and institutional constraints. The decision-making power of private charitable foundations, by contrast, is highly concentrated in the hands of donors and their management teams. Billionaires can use their wealth to choose what technology to fund, what research to support, and what problems to solve. Their influence may even extend to government policy and industrial development. This creates a paradox: the more technological progress may affect everyone, the more the power to decide the direction of technological development may be concentrated in the hands of a few.

Whether it is distributing money or distributing shares, what needs to be clarified first is that private charity and public finance have never been the same resource allocation mechanism.

Public finance is formed through taxation, and budgetary expenditure in principle requires legislative procedures, public discussion, and audit supervision. Government decisions may be slow and may be inefficient, but at least there are relatively clear channels of public accountability. Private charity is different. Donors can set the purpose of a fund according to their own value judgments, decide where the money goes and who benefits, and can also adjust or even terminate projects. Even if the funds are used for public undertakings, the decision-making process is not necessarily open to the public.

Political scholar Rob Reich has discussed this issue. He points out that the rich can convert private wealth into public influence through charitable donations without having to accept the constraints of democratic procedures like governments. The larger the scale of charity and the more far-reaching its impact, the more worthy of scrutiny is the ability of private wealth to shape public issues.

In the AI era, this contradiction has been further amplified. When a company like Anthropic participates in deciding how the losses caused by technological change should be compensated, to what extent can it represent the affected workers? Musk's proposal to give children SpaceX stock also attempts to let the public share in the gains from corporate growth. But for those who do not receive stock, and for those who may lose their jobs and income due to technological change, how can such an arrangement be called a safeguard?

I think one of the keys to solving these problems lies in who ultimately owns the productivity gains brought by AI. If the new gains are mainly converted into corporate profits and shareholder returns, then taking out a portion of the wealth afterward through charitable donations cannot solve the problem of how the whole society shares the fruits of technological progress. Donations can help some people, but they cannot replace labor protections, the tax system, and the social welfare system, still less can they change the ownership structure that has already formed in the process of wealth creation.

Effective altruism emphasizes rational calculation and impact evaluation to make limited resources produce the greatest social benefit. But when technology companies are both the drivers of change and participants in the related risks, donors and beneficiaries do not necessarily share the same interests. Simply calculating which option can help more people may not answer what is fair.

For example, a supermarket introduces an AI checkout system, reducing manual checkout positions; operating costs fall, and profits rise accordingly. The company then sets aside a sum of money to fund vocational training for the displaced employees. From the perspective of charitable impact, this money may help some people find new jobs, and may even help more people than direct subsidies. But for the employees who lost their jobs, what they may need is income security during unemployment, opportunities to transfer to other roles, and also the right to participate in the company's technological transformation and job adjustments. A company can calculate the input-output of a training program, but it cannot, on the basis of that calculation alone, decide for employees what kind of compensation is reasonable.

This reminds me that a couple of days ago, Ethereum co-founder Vitalik Buterin, in the podcast program "Silicon Valley 101," also talked about technological development, security, and the decentralization of power. As one of the most influential figures in the cryptocurrency field, Vitalik has long focused on decentralized technology and its social value. He proposed d/acc, that is, defensive accelerationism, which advocates continuing to push technological progress while prioritizing the development of technologies that can enhance society's defensive capabilities, so that more people have tools to protect themselves, rather than entrusting security entirely to a few centralized institutions.

He took epidemic prevention as an example. If there is concern that AI could help people create dangerous viruses, one approach is to restrict the most powerful models and prevent them from providing relevant information; another approach is to develop virus detection, air filtration, and vaccine research capabilities, so that society can detect risks faster and respond more effectively when they arise. The former approach tries to control dangerous capabilities, while the latter hopes to put defensive capabilities in the hands of more people.

If this line of thinking is extended to wealth distribution, what ordinary people need is not only compensation after technological progress, but also ownership to share in the gains from technological progress, and the right to participate in deciding the direction of technological development.

This article comes from the WeChat public account "Dongsi Shitiao Capital" (ID: DsstCapital), author: Wei Xianghui, published by 36Kr with authorization.