MisAnthropic letter questions Anthropic’s security ahead of billion-dollar IPO
What is the MisAnthropic Statement?
A public letter published on November 18, 2025, one day before Anthropic’s planned Nasdaq debut, criticizes the company’s reputation as a leader in artificial intelligence security. The signatories argue that the company uses this image to attract investors while weakening previous commitments and advocating for rules that would benefit it.
Who signed the letter?
The list includes experts from various fields and figures from the technology sector:
- Steven Pinker, Harvard psychologist
- Laurence Tribe, jurist
- Oliver Hart, Nobel laureate in economics
- Sheldon Glashow, Nobel laureate in physics
- Judea Pearl, winner of the Turing Award
- Fields Medalists
- Thomas Wolf, co-founder of Hugging Face
- Igor Babuschkin, co-founder of xAI
- Guillaume Verdon (Beff Jezos), advocate of accelerationism
- Joanne Harris, writer
- Mary Rasenberger, from the Authors Guild
Why is security at the heart of the dispute?
Anthropic was founded in 2021 by former OpenAI employees, including brothers Dario and Daniela Amodei, following disagreements over the direction of Sam Altman’s company. Since then, security has become a cornerstone of its brand. The company established itself as a public benefit corporation, created the Long-Term Benefit Trust, and adopted a responsible scaling policy in 2023, committing not to advance beyond certain risk limits without adequate protections.
The change in February 2026
In February 2026, Anthropic revised this policy. The new text abandoned the unilateral commitment not to proceed without specific protections and began to consider the company’s position in the race for superintelligence and the assessment of catastrophic risks. The revision also provides for public risk reports every three to six months and the obligation to match or exceed competitors’ security efforts.
Critics argue that the change conditioned any slowdown on the company’s position in the competition, reversing the original promise to prioritize security.
The dispute over AI laws
The Center for AI Safety (CAIS) accuses Anthropic of resisting legislative proposals such as SB 1047 (California), the RAISE Act (New York), and the AEGIS Act, while supporting rules it can already comply with, such as SB 53, also from California. CAIS believes the company is seeking to shape regulation in its favor.
A heterogeneous coalition
What stands out is the diversity of the signatories. Advocates of deceleration, such as members of the Future of Life Institute, stand alongside accelerationists like Verdon and George Hotz. They all agree that there is a contradiction between the discourse of security and the pursuit of increasingly powerful systems.
Nate Soares’ view
If I’m right, there will be no survivors.
Nate Soares, president of the Machine Intelligence Research Institute and co-author of “If Anyone Builds It, Everyone Dies,” warns that the race for superintelligence poses an existential risk, regardless of who wins. He proposes that the United States and China establish international oversight over the most advanced chips to prevent any country or company from advancing in isolation. Soares compares the situation to a bus heading toward a cliff, arguing that stopping the vehicle should be a priority before discussing probabilities.
Criticisms of the letter
Some experts considered the tone of the letter excessively harsh. Others noted that organizations agreed privately but preferred not to sign for fear of harming relationships or funding. The presence of competitors, such as Babuschkin (xAI), raises questions about possible conflicting interests. The letter does not present a concrete request to investors nor does it alter the IPO schedule.
Impact on the IPO
Anthropic’s Nasdaq debut is scheduled for November 19, 2025, under the ticker ANT. The maximum valuation of the offering could reach around $1.4 trillion on a fully diluted basis. The company projects revenue of $33.2 billion in the first nine months of 2026 and has structured its control to protect its mission from market pressure. By questioning the company’s security reputation, the signatories aim to make new shareholders assess whether the company’s image matches the risks involved.
Frequently Asked Questions
- What is the MisAnthropic Statement? A public letter that questions Anthropic’s security reputation and accuses the company of weakening commitments and advocating for advantageous rules.
- Who signed the letter? Among the signatories are Steven Pinker, Laurence Tribe, Oliver Hart, Sheldon Glashow, Judea Pearl, Thomas Wolf, Igor Babuschkin, Max Tegmark, and Nate Soares, among others.
- What changed in Anthropic’s security policy? In February 2026, the company replaced the unilateral commitment not to advance without specific protections with a policy that considers its position in the race and the assessment of risks.
- Does Anthropic support AI safety laws? The company supported SB 53 in California, but CAIS accuses it of resisting other proposals, such as SB 1047 and the RAISE Act.
- Did the letter change the IPO schedule? No. The Nasdaq debut is still scheduled for November 19, 2025, under the ticker ANT.








