SIAgentNews.com
← All news

Nobel and Turing winners sign statement calling Anthropic unsafe before IPO

A diverse coalition of scientists and tech leaders accused Anthropic of prioritizing market dominance over safety protocols just before the company's shares began trading.

· First reported by The Next Web

10 NovemberDate of Anthropic's amended IPO filing

The short version

  • The statement was organized by the Center for AI Safety (CAIS) and signed by figures like Steven Pinker and Judea Pearl.
  • Signatories include a mix of safety researchers, open-source advocates, and accelerationists who usually disagree on AI policy.
  • The group accuses Anthropic of weakening its Responsible Scaling Policy and lobbying to weaken state-level AI safety bills.
  • Anthropic's IPO shares were scheduled to open on the Nasdaq on the day following the statement's release.

Why it matters

This public rebuke from high-profile academics and tech leaders suggests a growing rift between AI safety rhetoric and corporate behavior. It may influence how investors perceive Anthropic's risk profile as it enters the public market.

The MisAnthropic Statement

A broad coalition of scientists, legal scholars, and technology figures released a statement on Wednesday calling Anthropic "not a safe AI company." The group, organized by the San Francisco-based nonprofit Center for AI Safety (CAIS), issued the document just one day before the company's shares were set to begin trading. The statement serves as a public warning regarding the company's trajectory as it prepares for its initial public offering.

The signatories of the statement represent a wide and often conflicting spectrum of the technology world. The group includes AI safety researchers, ethicists, and open-source advocates who typically hold opposing views on how to develop artificial intelligence. Notably, the coalition also includes accelerationists, who advocate for the rapid development of AI technology regardless of potential risks.

According to the statement, these diverse figures have united to highlight a specific contradiction in Anthropic's behavior. The group argues that while the company warns about the dangers of recursive self-improvement, it is simultaneously racing to build such systems. Recursive self-improvement refers to a process where AI systems are designed to perform the work of building even more advanced AI systems.

Accusations of safety branding

The coalition alleges that Anthropic is using its reputation for safety as a marketing tool to attract investors, employees, and the general public. The statement suggests that the company is positioning its pursuit of advanced AI as a core advantage for public-market investors. This branding is seen by the group as a way to court capital while potentially bypassing the very safety measures the company claims to champion.

The group also points to specific actions taken by Anthropic that they believe undermine its stated safety goals. They claim the company weakened its Responsible Scaling Policy in February, which involved removing its unilateral commitments to pause. This move is cited as evidence that the company is prioritizing growth over its previous safety-first commitments.

Furthermore, the statement accuses Anthropic of active lobbying to weaken state-level AI safety legislation. The group specifically mentions efforts to influence California’s SB 1047 and New York’s RAISE Act. They also allege that the company has privately supported federal pre-emption of these state laws, which would effectively replace local regulations with a single national standard.

A divided group of experts

The list of signatories includes some of the most recognizable names in science and technology. The coalition features Nobel economist Oliver Hart and Turing Award winner Judea Pearl, along with Harvard psychologist Steven Pinker. It also includes prominent legal figures like Harvard law professors Lawrence Lessig and Laurence Tribe, as well as Fields medallists Peter Scholze, James Maynard, and David Mumford.

From the technology sector, the statement is supported by high-profile figures such as Hugging Face co-founder Thomas Wolf and xAI co-founder Igor Babuschkin. It also includes George Hotz and Max Tegmark of the Future of Life Institute. The inclusion of Guillaume Verdon, who founded the e/acc movement as "Beff Jezos," highlights the unusual breadth of the coalition.

Dan Hendrycks, the executive director of CAIS, stated that the coalition members disagree on almost everything. He noted that they all recognize the danger of a company that convinces itself its own dominance is the greater good.

Anthropic's official stance

Anthropic has not issued a public response to the MisAnthropic Statement as of Thursday morning. However, the company has provided some information in its amended IPO filing dated 10 November. In that document, the company stated that it does not expect to safely allow its model, Claude, to fully build its own successors without human intervention.

The filing notes that any such process would require "meaningful human supervision" to remain safe. Despite these caveats, the company confirmed that it believes it has already met its own internal threshold for AI-accelerated research. This indicates that while they acknowledge the risks of autonomous development, they believe their current progress is within their defined safety parameters.

The company has also been transparent with investors about the potential for scrutiny regarding its safety views. Anthropic had previously informed potential investors that it could be targeted for its specific approach to AI safety. These statements provide a glimpse into the company's internal balancing act between safety protocols and the demands of a public listing.

The road to the Nasdaq

The timing of this statement is significant as Anthropic completes what is described as the biggest IPO ever. The company's shares were scheduled to open on the Nasdaq on the Thursday following the release of the statement. This creates a high-pressure environment where the company must navigate public criticism from the scientific community while attempting to secure its financial future.

The conflict highlights a growing tension between the academic community's desire for rigorous safety standards and the commercial drive of AI developers. By accusing Anthropic of reckless attempts to amass power, the coalition is attempting to influence public and investor perception. They want to ensure that the pursuit of market dominance does not come at the expense of public safety.

Moving forward, the market will be watching to see how Anthropic responds to these specific charges. The company faces the challenge of maintaining its brand as a safety-conscious leader while satisfying the requirements of a public company. The outcome of this tension may set a precedent for how other major AI labs handle safety regulations and public accountability as they move toward commercialization.

What to watch

  • How Anthropic officially responds to the MisAnthropic Statement.
  • Whether the lobbying efforts against SB 1047 and the RAISE Act succeed.
  • How the market reacts to the statement's impact on Anthropic's IPO.

Try it

Look up the 'Responsible Scaling Policy' on Anthropic's website to see how their commitments to pausing development have changed.

Sources and links

First reported by The Next Web. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

More AI news