Coalition challenges Anthropic safety claims ahead of IPO
Reported by Yahoo Finance UK, news.lavx.hu, FourWeekMBA and 4 more outlets7 sources ↓
A coalition of AI researchers, founders, legal scholars and safety advocates published the MisAnthropic Statement, arguing that Anthropic’s pursuit of AI capable of building better AI conflicts with its safety warnings and branding.
Key points
- The statement accuses Anthropic of pursuing recursively self-improving AI while warning of its risks
- Signers say Anthropic uses its safety reputation to attract investors, employees and public support
- The coalition also criticizes changes to Anthropic’s safety pledges and its positions on regulation
- Anthropic’s planned Nasdaq debut is the context for the public appeal
Sources
- Yahoo Finance UKRead ↗"Anthropic is Not a Safe AI Company" - Broad Coalition of Scientists, Legal Scholars, and Tech Leaders Challenge Anthropic’s Safety Branding Ahead of IPO
- news.lavx.huRead ↗Coalition challenges Anthropic safety claims ahead of IPO
- FourWeekMBARead ↗89 Sign MisAnthropic Statement Hours Before Anthropic IPO
- Hacker NewsRead ↗"Anthropic Is Not a Safe AI Company" – Coalition Challenges Anthropic Safety
- scanx.tradeRead ↗Coalition accuses Anthropic of misleading safety branding ahead of IPO
Details
The Center for AI Safety announced the statement on November 18. The signers say Anthropic’s warnings about AI that could build more capable AI conflict with its efforts to gain an advantage by developing the technology. They call the company unsafe and urge the public to reject what they describe as a power grab under a safety rationale.
The statement also accuses Anthropic of using its safety reputation to win support and of weakening its pledges while lobbying for rules that serve its interests. CAIS cites California’s SB 1047, New York’s RAISE Act and a federal proposal called the AEGIS Act. The statement leaves the choice of legal or enforcement measures open.
Anthropic’s February revision to its voluntary Responsible Scaling Policy removed its general unilateral pause commitment while introducing public safety roadmaps and recurring risk reports. The company retained conditional commitments to delay some activities, including when it had clear evidence it was ahead and had not contained a catastrophic risk. Anthropic said uncertainty over when models cross capability thresholds made an industry-wide case for action difficult.
Anthropic’s November 17 amended registration statement says each Claude generation helps speed development of the next and identifies AI-assisted research as a safety risk. The company estimated that Claude led 26% of its model research and development work in August, compared with under 1% in February; it estimated no measured work was fully autonomous. These are Anthropic’s internal estimates.
Written from coverage as of
Images
- kingy.ai
- kingy.ai
Stocks in this story
Change since open: 20 Nov to 21 Nov 2026
Last refresh