AI companies often talk about benefiting society. On 18 October, Anthropic described a more concrete structure: a new fund will hold shares in the company and may use their value for purposes such as grants to nonprofits or direct payments to people.
According to Anthropic, 5 percent of its shares will be committed to the fund, subject to stockholder approval. The board may add another 2 percentage points each time the company’s value rises by 50 percent above a defined baseline. Including earlier pledges by founders and employees, the company says roughly 20 percent of its shares will be dedicated to the public good.
The hopeful part is the link between growth and responsibility: if the company becomes more valuable, the resources set aside for public benefit can grow as well. Anthropic mentions support for nonprofits, research and possible direct distributions to groups affected as AI changes the labour market.
But this is still a promise made by the company itself. Stockholders must approve the initial transfer, future additions are at the board’s discretion, and details about governance and valuation thresholds will come later. What the fund actually does will therefore matter more than the headline percentage.
If the structure works as intended, it could provide an interesting model: not only donating after profits have been made, but tying a slice of ownership itself to the public good.