On paper, Anthropic’s pledge is worth roughly $100 billion. Elon Musk took less than a day to respond with an offer of his own, valued at about $10.5 billion.
On Sunday, the maker of the Claude chatbot announced it would put 5% of its shares into a new Public Benefit Fund. A day later, on Monday, Musk promised that every child in the United States would get one SpaceX share from him next year.
A fund that’s set up as a company
How the fund is built is the part worth watching. Anthropic made it a wholly owned LLC instead of a nonprofit, and the company says that setup gives it room to switch between different forms of giving.
“The Anthropic Public Benefit Fund will have increased flexibility and scale, enabling us to pursue our mission in new ways: for example, we might explore large donations to traditional nonprofits or direct distributions to specific populations,” Anthropic said. It has already set aside $200 million to study those options.
Each time the company’s value climbs 50% past a baseline, the board can move another 2% of shares into the fund. That baseline hasn’t been made public. The fund’s equity can’t be spent on lobbying or political donations.
Several details are still unknown. Anthropic hasn’t explained how the fund will be taxed or what it will report publicly, both of which a traditional charity has to address. The company says the governance structure and valuation thresholds will be published in the coming months.
The $100 billion figure depends on an IPO
Anthropic is getting ready for an IPO, where it would sell shares to the public on the stock market. That could happen as early as November, with a reported valuation of around $2 trillion.
At that valuation, a 5% stake works out to about $100 billion. That number is the value of the shares. There’s no $100 billion pile of cash sitting in a bank account.
The fund isn’t the only thing Anthropic is counting. Its co-founders have pledged 80% of their equity to charity, and the company matches employee donations. Put together, Anthropic says about 20% of its shares will go toward the public good.
David Sacks, an investor and former White House AI czar who has criticized Anthropic before, wasn’t persuaded. “You’ve put 20% of your equity in an LCC. Whether it serves the public good remains to be seen,” Sacks said in a post on X.
Musk’s counteroffer runs through Trump Accounts
Musk’s first reply came Sunday night, when he suggested “giving the shares directly to the public” in place of sending them through a fund. His pledge to children arrived early Monday.
[[IMG:1]]
“Next year, I will give one of my SpaceX shares to every kid in America,” Musk wrote on X. “This will be the first of many.” He pitched it as a step toward Universal High Income, which he offers as an alternative to Universal Basic Income.
With roughly 72 million children in the U.S. and SpaceX shares trading at about $147 apiece, the pledge totals around $10.5 billion. According to an SEC filing, that’s a little over 1% of the 6.42 billion shares Musk owned as of June 30. It’s also about 1% of his estimated $908.5 billion fortune, which has topped $1 trillion several times this year.
The shares will likely reach kids through Trump Accounts, the tax-advantaged savings accounts for children. The Treasury Department recently began allowing companies and charities to donate individual company shares to them, and donated stock is generally held for five years. Musk all but confirmed this in a reply on X: “Thanks for helping enable this.”
[[IMG:2]]
The doom talk and the safety bill
The donation comes from a company that warns about AI risk as loudly as anyone while releasing products just as quickly. In January, CEO Dario Amodei wrote that “we are considerably closer to real danger in 2026,” yet Anthropic continues to put out the most powerful AI models available.
In April, OpenAI CEO Sam Altman accused Anthropic of using fear-based marketing to sell Claude Mythos. Now, ahead of its market debut, the company is handing out part of itself.
Americans are uneasy about how fast things are moving. A late-September AP-NORC poll found 64% of U.S. adults think AI is developing too quickly, compared with 8% who say it’s too slow.
Two days before the fund announcement, Anthropic quietly passed lawmakers a draft AI safety bill of its own. Under it, the government could stop the release of a model that poses a foreseeable, imminent risk of killing or seriously injuring more than 50 people, or of causing more than $1 billion in damage. The bill lists three routes to that kind of harm: a bioweapon, an AI escaping human control, or a cyberattack.
Any block would need to pass through the Commerce Department and get a federal court order. Developers would also have to publish safety frameworks and transparency reports for new models.
According to Anthropic, the bill is based on its Advanced AI Framework, which it has been pitching to lawmakers since June. That same month, Amodei called for binding rules on frontier models, the term for the most powerful AI systems. The IPO could arrive as soon as November, and the fund’s rules have yet to be published.