In the same week that Anthropic told investors its quarterly revenue had jumped 45% to $16.8bn, the company announced that it would give away part of itself. On Sunday it said it would commit 5% of its shares to a Public Benefit Fund, with a further 2% possible at the board's discretion. The fund could back nonprofit organisations and efforts to soften AI's impact on the jobs market, though none of its shares may be used for lobbying or political donations, in the United States or abroad.

The timing is hard to ignore. Anthropic is laying the groundwork for a stock market listing that could value it at more than $1tn. Its filing reportedly acknowledges that AI could pose catastrophic or existential risks to humanity, a striking admission for a company asking investors to buy in. Critics may read the fund as reputation management; supporters may see a rare attempt to build social responsibility into a company's ownership before it goes public.

The pressure behind the gesture is real. The International Monetary Fund estimated in 2024 that almost 40% of jobs worldwide will be affected by AI, rising to as much as 60% in advanced economies. Technology firms from Silicon Valley to Asia have already cut thousands of posts after investing in the technology. Meanwhile, Jacob Coxon, a 27-year-old researcher who left Anthropic, said people building AI were genuinely frightened by its speed.

Elon Musk, himself a rival in the AI race, replied that shares should simply go directly to the public, adding that he would do so himself.