Will Anthropic Keep AI Safety First After Its IPO?

Anthropic says it will keep safety ahead of profit through three safeguards: public benefit corporation status, a single founder share with 50.1% of the voting power, and a Long-Term Benefit Trust that shapes the board. Whether those hold up is the real question around Anthropic AI safety as the company prepares to list, with trading expected as soon as the week of November 16, according to CNN.
Anthropic Is a Public Benefit Corporation, Not a Typical Public Company
Anthropic is incorporated as a Delaware public benefit corporation. Its amended SEC filing says the company must balance stockholders’ financial interests, the interests of people affected by its conduct, and its stated public benefit, and calls the structure consistent with its commitment to AI safety. OpenAI is also a PBC, but Dow Jones notes it’s controlled by a charity, not a trust.
Founders Hold 50.1% of Votes Through a Single Share
A single special share held by a company owned by Anthropic’s seven co-founders carries 50.1% of the voting power, VC Corner reports. Ordinary shareholders get one vote per share, but they can’t outvote the founders.
A Trust Elects the Board and Can Block Major Moves
The Long-Term Benefit Trust is meant to be an independent body of trustees with no financial stake in the company and expertise in AI safety, national security and public policy, per the SEC filing. Dow Jones says it holds enormous sway: it can shape the board and block major corporate actions.
The details are still murky. The trust’s governing documents haven’t been published, and shareholders could only change its powers if a very large supermajority agreed. One law professor told Dow Jones that, without major changes, public shareholders will have very few governance rights.
Anthropic Is Also Putting 5% of Shares Into a Public Benefit Fund
Anthropic has also unveiled a plan to place 5% of its shares in a fund that could pay for donations to nonprofits or other mission-related work, Dow Jones reports. It’s an attempt to show the mission isn’t only a legal structure.
The Real Test Comes When Profit and Safety Pull Apart
The company isn’t hiding the tension. Its prospectus devotes 93 pages to risk factors, which we covered in our earlier look at the leaked filing. The money side is equally big: revenue rose to $16.3 billion in six months, but net losses topped $80 billion, as detailed in our breakdown of the S-1.
So will Anthropic keep AI safety first after its IPO? The structure gives it the tools to try. The strain will show if earnings disappoint, a safety problem appears, or the board, founders and trustees disagree.