Anthropic recognises the risks to humanity – but is its governance really up to the job?
Reported by The Guardian1 source ↓
Anthropic’s governance gives its seven founders and a trust central roles in company decisions, with ordinary shareholders holding little power. The Guardian questions whether the untested structure can provide effective oversight of AI safety.
Key points
- The seven founders collectively hold 50.1% of Anthropic’s ultimate voting power
- The Long-Term Benefit Trust nominates four of the company’s seven directors
- The Guardian reports Anthropic’s listing document cites potential catastrophic or existential AI risks
- How the trust would challenge the founders over a safety dispute is unclear
Details
Anthropic’s governance divides power between its seven founders and the Long-Term Benefit Trust, rather than giving ordinary shareholders the central role. The founders collectively hold one vote carrying 50.1% of ultimate voting power, while the trust nominates four of the seven directors. The trust is intended to balance financial returns with benefit to humanity.
The Guardian describes the arrangement as an untested experiment. It says the structure may create a check on founder power, but questions whether trustees would publicly challenge the founders if they disagreed over AI safety, given the founders’ voting control.
The article cites investment manager Stuart Riddick’s point that agreement with a mission does not ensure agreement on how directors interpret it. It also notes Microsoft AI chief Mustafa Suleyman has criticised Anthropic’s views on model welfare, including whether AI systems should be trained to behave as if they have rights or consciousness.
Written from coverage as of
Images
- The Guardian
Stocks in this story
Change since open: 20 Nov to 21 Nov 2026
Last refresh