Anthropic, the company behind the Claude AI models, says it will put 5% of its shares into a new Anthropic Public Benefit Fund, subject to approval by its existing shareholders. The board may add a further 2% for every 50% rise in the company's value above a baseline that has not yet been disclosed.

Anthropic gave no rand or dollar figure, but at its last reported valuation in May, 5% would be worth about $ 48 billion, roughly R811 billion, on paper.

The fund will be a limited liability company owned by Anthropic rather than a traditional charity, which the company says gives it flexibility to make large donations to nonprofits or direct payments to specific groups of people. One stated focus is AI's effect on jobs. Anthropic has already committed $ 200 million, about R3.4 billion, to research on that question, and said that given how fast AI is moving, bolder approaches may be needed.

It says none of the fund's equity will be used for lobbying or political donations in the United States or elsewhere. Counting earlier pledges by its founders and an employee donation-matching programme, Anthropic says roughly 20% of its shares are now set aside for public benefit.

What is missing matters as much as what was announced. The fund's governance, who decides what it pays for, and the valuation baseline that would trigger extra shares will only be published in the coming months. The shares also only turn into money that can be given away if they can be sold, and Anthropic is reportedly preparing to list on the stock market as soon as November, having told investors its third-quarter revenue was $ 16.8 billion, about R284 billion.

The announcement comes amid growing public unease about AI, which chief executive Dario Amodei has called a crisis of trust fundamentally.

TechCrunch's Anthony Ha noted that Anthropic has not yet said who will decide which causes are worth funding. CNBC quoted Lise Buyer of IPO advisory firm Class V Group, who said the timing is both necessary preparation for a listing and beneficial from an image perspective, and reported Elon Musk's response on X recommending the shares be given directly to the public. Reuters confirmed that the additional 2% top-ups are at the board's discretion.

Within a few weeks, AI companies have tried three different responses to public worry. OpenAI held back its GPT-6.1 Astra model over safety concerns. It also began watermarking ChatGPT text in the EU, though only because the law required it. And now Anthropic is pledging money.

Of the three, money is the largest in scale and the least defined. A pledge of shares, decided and governed by the company, with key details still to come and timed ahead of a listing, deserves a measured reading: it could fund real help for workers affected by AI, or it could mostly signal good intentions. The test will be whether the fund is run independently, reports openly on what it spends and where, and reaches people outside the US.

That last point matters here. South Africa's call centres and back offices are among the jobs most exposed to AI, as we covered in our piece on the local BPO sector, and nothing yet says whether money like this will reach them. Disclosure: The Open Letter uses Claude in its production work, and this post was drafted with it.

You might also like our piece on the AI slowdown calls filling feeds, Character.AI filing for bankruptcy, and Claude becoming a full office suite.

Get more SA tech and business news and subscribe to The Open Letter.