Anthropic proposes a public-benefit fund holding 5% of its shares

Anthropic announced on 18 October that it plans to commit 5% of its shares to a new Public Benefit Fund, subject to stockholder approval. Further additions of 2% would be at the board’s discretion for every 50% increase in company value above a set baseline.[1]

The proposed vehicle is a wholly owned limited liability company, with scope for nonprofit donations or direct distributions to particular populations. Anthropic says it will publish governance and valuation-milestone details in the coming months; the announcement gives neither the baseline value nor a timetable for payouts.[1]

A research programme already asks who should benefit

There is a concrete starting point for understanding the choices. In July, Anthropic committed $200 million to an Economic Futures Research Fund examining responses to AI’s economic effects. Its agenda invites universities, research institutes and nonprofits with experience running field studies to test interventions. Most individual projects are expected to receive $5 million to $30 million, allowing experiments large enough to study changes in workplaces and communities.[2]

The agenda spans quite different kinds of help. One strand examines training and job transitions, including combinations of retraining and income support. Another asks about unemployment-insurance reforms and support for basic needs. A third explores ways for workers and communities to hold a financial stake in economic gains, through possibilities such as capital accounts or equity-sharing arrangements.[2]

These are research questions in a separate funding programme. They illustrate why deciding to share AI’s gains leaves another substantial decision: whether to help someone change jobs, replace lost income, or acquire an asset that can grow in value.[2]

Funding and company governance do different jobs

Anthropic has also used a different mechanism to pursue its public purpose. When it introduced the Long-Term Benefit Trust in 2023, it described independent, financially disinterested trustees with rights to select and remove members of the company’s board. That design sought to bring the interests of people affected by AI into corporate governance. The trustees’ role was defined through board-selection rights, rather than through a programme for distributing money to beneficiaries.[3]

Governance arrangements concern who can influence a company’s decisions. A fund concerns which resources can be directed towards a purpose and how they reach recipients. An ownership allocation can establish a financial resource without, by itself, specifying the public programme that will use it.[3][1]

The forthcoming fund rules will determine how a growing corporate stake becomes support people can use: who can receive it, who decides, and when distributions happen.[1]

Report an error or suggest a correction