Anthropic’s 5% public benefit fund is an IPO move with a real idea inside: sharing AI’s gains with the workers it displaces

Anthropic announced on Sunday that it plans to commit 5% of its shares to a new Public Benefit Fund, pending shareholder approval. The board may add 2 percentage points for every 50% increase in the company’s value above a baseline, though it has not disclosed that baseline. According to the company, the fund could make large donations to traditional nonprofits, make direct payments to specific groups of people, or pursue ‘bolder approaches’ to studying and addressing AI-driven job disruption and displacement. Anthropic had already committed $200 million to studying AI’s effects, but says ‘given the pace of AI progress, bolder approaches may be needed.’

The structure matters. The fund will be a wholly owned limited liability company, not a 501(c)(3) nonprofit, which Anthropic says will let it switch between forms of giving. It also says the equity held by the LLC will not be used for lobbying or political donations, in the US or anywhere else. Details on governance and the valuation milestones are promised ‘in the coming months,’ with progress to be reported in a future public benefit report.

The context is hard to ignore. According to CNBC, Anthropic told investors its third-quarter revenue is $16.8 billion, ahead of an IPO reportedly planned for November. Lise Buyer of IPO advisory firm Class V Group told CNBC the announcement is partly required preparation for going public and partly ‘beneficial from an image perspective,’ and is also meant ‘to burnish their reputation at a time when AI is coming under significant scrutiny.’ Public anxiety is real. A Pew Research Center report from August found that 71% of US adults expect AI to mean fewer jobs over the next two years. Elon Musk replied on X that he ‘recommend[s] giving the shares directly to the public.’

Skepticism is fair. A fund with no disclosed baseline, unclear governance and broad discretion is so far a commitment of intent, not a program. An LLC gives flexibility, but it also gives less outside oversight than a regulated charity. And because the stake grows with valuation, the fund’s size depends on the same success that may cause the disruption it is meant to ease. Some will see that as a well-designed hedge. Others will see a conflict of interest.

Our reading: we have argued that Anthropic’s IPO turns capital markets into a third channel of AI governance, after litigation and regulation. This fund goes further. It writes a version of redistribution into the cap table, and audited financial disclosure will eventually show whether the money actually reaches displaced workers or just funds reports. That is the test to watch. The short-term transition is hard and uneven, and back-office workers are the most exposed. A long-term future of abundance only works if the gains are shared, and a frontier lab tying part of its equity to that goal sets a precedent rivals will be pressed to match. The motive may be partly cosmetic. Whether it matters will depend on governance, and so far Anthropic has only promised to explain it later.