Anthropic Places AI Safety in Hands of Handpicked Trustees as Company Goes Public

Anthropic is going public with a governance structure that hands significant power over its most powerful AI models to a small group of financially independent trustees, betting that external oversight can temper the profit motives of rapid growth. The company's IPO filing, disclosed October 26, 2026, reveals that a Long-Term Benefit Trust—holding no equity and compensated at $400,000 per trustee annually—will elect a majority of Anthropic's board within four years, even as seven co-founders maintain voting control through a golden share mechanism and early investors like Amazon and Google hold only nonvoting stock.