Anthropic proposes putting 5% of its shares into a fund to address AI's impact
The company behind Claude plans to finance grants and possible direct assistance through an entity it owns. The proposal requires shareholder approval and comes amid concern about artificial intelligence's effects on employment.
Anthropic wants to link part of its growth to the social costs that artificial intelligence may bring. The company behind Claude announced on Sunday, October 18, that it proposes allocating 5% of its shares to a new public benefit fund that could support organizations and explore direct distributions to particular population groups. The transaction is subject to shareholder approval.
The announcement raises a question at the heart of the AI debate: how to share some of the wealth generated by a technology capable of transforming work. The company acknowledges that its advances could create substantial economic value while also causing instability for many workers. It has not yet set out a specific program identifying recipients, payments or a timetable for assistance.
The Anthropic Public Benefit Fund will be an entity within the group. Its stake could increase if the company's valuation rises, although that mechanism is not automatic: additional share allocations will be at the board's discretion.
An AP-NORC poll published on October 8 offers a measure of that unease: 64% of U.S. adults think AI is developing too quickly. The survey of 2,140 adults, conducted September 24–28, records opinions about the technology, rather than estimating how many jobs will disappear.
A stake that could grow with the company
Under the announced plan, for every 50% increase in Anthropic's value above a specified benchmark, the board could allocate another 2% of shares to the fund. The company still needs to explain that benchmark and the valuation milestones that would allow further contributions.
The distinction between an equity stake and a cash transfer is important in assessing the promise. The 5% represents shares; it does not mean that the same proportion of revenue or profit is already being distributed to workers. Its economic value will depend on the company and on how the fund's resources are realized.
Reuters places the initiative at a time when Anthropic is preparing to go public and facing growing questions about its systems' safety and social consequences. The announcement commits equity, but does not yet make it possible to measure how much money will ultimately reach each activity or what it will achieve.
Employment was already the focus of a separate program
Employment is one possible use of the fund, rather than its only purpose. The company had already committed $200 million to its Economic Futures Research Fund, whose research agenda it published in July. That program seeks to finance external research and trials of interventions to help workers, businesses and governments prepare for AI-related economic changes.
Its priorities include studying how to introduce these tools in workplaces, improve training and career transitions, and adapt support for people who lose their jobs. It also aims to examine ways for workers to share in the gains before any potential disruption arrives.
The agenda envisages experiments and program evaluations, including initiatives for learning a profession when systems take over tasks that once provided an entry point into an occupation. These are research questions: their inclusion does not establish that a proven solution already exists or that every profession will change in the same way.
The new equity fund broadens the options Anthropic wants to have available. The company presents it as a more flexible instrument than research alone, potentially supporting large-scale grants or interventions aimed at specific populations. Those populations have yet to be selected.
What has been promised, and what remains unexplained
The entity will be set up as a limited liability company wholly owned by Anthropic. The company chose that structure, rather than a conventional U.S. charitable organization, so that it could combine different ways of providing support.
Anthropic says the fund's shares will not be used for lobbying or political donations, either in the United States or abroad. That commitment is part of the announced design; compliance and the results of its activities will need to be assessed once the rules are known and operations begin.
The company promises to provide more information about structure and governance in the coming months. Unresolved questions include how allocations will be decided and how committed resources will translate into verifiable benefits. Commercial research, social initiatives and management decisions do not necessarily serve identical purposes.
A response that is already prompting other promises
The announcement drew an immediate reaction from Elon Musk. According to Associated Press, the entrepreneur argued that shares should be given directly to the public and promised to give every U.S. minor one SpaceX share next year. He did not explain how that would work. His proposal is separate from Anthropic's fund and is not a distribution that has already taken place.
Anthropic's founders had pledged to donate 80% of their own holdings. The company also runs an employee share-donation matching program, in which roughly half its staff participate. AP reports that, taken together with the new proposal, these initiatives would dedicate about 20% of the company's shares to public benefit purposes. That percentage covers the combined commitments, rather than just the newly announced fund.
The next test will be turning the announcement into rules and action. Contributing shares can tie social funding to a company's success, but the actual impact will depend on decisions still to come: approval, fund governance, available resources and the eventual recipients of assistance.
