Anthropic Allocates 5% Stake to Public Fund; Musk Announces SpaceX Equity Gift

This fund follows a draft AI safety law that Anthropic presented to lawmakers a few days earlier. The draft would allow the government to block AI models that could cause more than 50 deaths or injuries, or lead to losses of more than $1 billion.

Anthropic said: “The Anthropic Public Benefit Fund will have more freedom and resources to advance our mission in new ways. For example, we may consider making large donations to traditional nonprofit organizations or giving money directly to specific groups.” The company said it has already committed $200 million to research this effort.

Americans are worried that AI is developing too quickly. An AP-NORC poll conducted at the end of September found that 64% of U.S. adults believe the technology is advancing too fast, while 8% think it is moving too slowly.

The fund is set up as a limited liability company (LLC) fully owned by Anthropic, rather than as a nonprofit organization. Anthropic says this structure makes it easier to switch between different types of donations. The fund's board can receive an additional 2% of the company's shares if Anthropic's value rises to 50% above a benchmark that has not been made public. The fund cannot use any of its shares for lobbying or political donations.

Elon Musk also wants attention. He has promoted “Universal High Income,” which is different from Universal Basic Income, and wrote on X that he would give every child in the United States a share of his SpaceX stock. “This will be the first of the firsts,” he added.

Anthropic's decision comes from a company whose CEO wrote in January, “By 2026, we will be quite close to real danger.” The company is known for warning about the possibility of AI causing disaster, even as it continues developing the technology at full speed and has launched some of the most powerful AI models on the market.

Two days before announcing the fund, Anthropic quietly sent lawmakers its own draft AI safety law. The draft says the government could block the release of an AI model if there is a foreseeable and immediate risk that it could cause more than 50 deaths or serious injuries, or more than $1 billion in losses. The risks could come from biological weapons, AI escaping human control, or cyberattacks.

Any such ban would need approval through an order from the Department of Commerce and an order from a federal court. Developers would also have to publish a security plan and a transparency report for new models. Anthropic said the draft law reflects its Advanced AI Framework, which it has been promoting to lawmakers since June. That same month, CEO Dario Amodei called for legally binding rules for frontier models, meaning the most powerful AI systems.

Some critics of Anthropic include investors and David Sacks, a former White House AI adviser known as the AI Czar. Responding to the company's commitments on X, Sacks wrote: “You put 20% of your equity into an LCC. Whether it truly serves the public interest remains to be seen.”

Anthropic said it would share details about how the fund is governed and the valuation limits in the coming months. The company could go public as early as November.