Anthropic wants seven co-founders to control 50.1% of the vote after its IPO

Anthropic is asking shareholders to approve a structure that would give Dario Amodei and six other co-founders a combined 50.1% of voting power ahead of a possible public offering. The Information describes the proposal as modelled on Palantir, where founder control persists even as economic ownership disperses.

This is more specific than Anthropic’s earlier plan to create supervoting shares. A collective majority means the founders could determine the outcome of contested shareholder votes without individually owning anything close to half the company. Amodei was previously reported to hold roughly 2% economically. The arrangement would sit alongside Anthropic’s Long-Term Benefit Trust, whose non-shareholder trustees can elect a majority of the board and are meant to protect the company’s public-benefit mission.

The two control mechanisms address different fears. Founder voting power resists pressure from public investors; the trust is supposed to resist both founders and investors if commercial incentives threaten the long-term benefit mission. But layering them also obscures accountability. If the founders and trustees disagree, investors need to know which body actually prevails, how trustees can be removed, and whether the founder bloc can amend the rules that constrain it.

The safety case is plausible: a frontier laboratory making decisions with large external consequences should not be forced into quarterly optimisation. The governance cost is equally real. Public shareholders would provide capital while possessing limited ability to replace leaders after strategic mistakes. Anthropic is therefore not merely designing an IPO; it is asking markets to finance a constitution whose most important checks have not yet been tested under pressure.