Anthropic just structured a Public Benefit Fund with 5% of company shares upfront, scaling up to ~20% total based on valuation milestones (every 50% increase unlocks another 2% at board discretion). This is separate from the co-founders' 80% personal equity pledge and the employee donation matching program they've been running since launch.
Technically interesting because it's a wholly owned entity designed to lock capital into public benefit work—not just vague CSR promises. The scaling mechanism ties public contribution directly to company success, which could set a precedent for AI labs balancing profit motives with alignment research funding.
Still unclear: what counts as "public benefit" in execution? Research grants? Open safety tooling? Governance experiments? The structure is clever but the devil's in the deployment details.
Technically interesting because it's a wholly owned entity designed to lock capital into public benefit work—not just vague CSR promises. The scaling mechanism ties public contribution directly to company success, which could set a precedent for AI labs balancing profit motives with alignment research funding.
Still unclear: what counts as "public benefit" in execution? Research grants? Open safety tooling? Governance experiments? The structure is clever but the devil's in the deployment details.