Skip to content

Leading figures condemn Anthropic for reckless race to recursive self-improvement

Published:
November 18, 2026
Author:
Future of Life Institute

Contents

Experts agree that the breakneck pursuit of recursive self-improvement (RSI) – AI that can build itself – will rapidly create systems we cannot control, and could cost millions of lives.

In response, a group of preeminent figures – including authors like Steven Pinker and Rutger Bregman, cofounders like Igor Babuschkin (xAI) and Thomas Wolf (Hugging Face), Nobel Laureates, Fields Medallists and more – have cosigned a statement decrying the company’s activities, in particular RSI, arguing that “We must curb Anthropic’s political influence and reject its reckless bid for technological dominance”.

Among the signatories were Future of Life Institute cofounders Max Tegmark (Chair) and Anthony Aguirre (President and CEO).

“When it comes to recursive self-improvement (RSI), supposedly “responsible” Anthropic is the most guilty of all the AI companies endangering everyone to enrich themselves,” Tegmark writes. “They wear two faces: In public, their executives warn it will create uncontrollable systems that could cause human extinction within a decade. In private, they are pedal-to-the-metal on RSI, driving the race and boasting to prospective investors that it is their “competitive advantage”. Ahead of this IPO, people should consider whether the ROI is worth risking the lives of their loved ones.”

“Most experts believe that if we allow AI systems to autonomously improve themselves or build their own successors, we will lose our ability to monitor, understand or control them, threatening billions of lives. In kickstarting and accelerating the pursuit of RSI, Anthropic reveals that its safety commitments are not worth the blogs they are posted on.” explains Aguirre. “The word “anthropic” literally means human-centric, but right now their name is perversely ironic. If Anthropic is serious about keeping humans safe, alive and in-charge, they should immediately stop the drive for RSI and push for binding legislation that requires their competitors to do the same.”

Some history of Anthropic and RSI:

  • Anthropic has spent years telling the world to slow down. In 2023, the company publicly warned about rapid AI progress, stipulating that “the use of advanced AI in AI research” could accelerate catastrophic risks. In September 2026, Dario Amodei sounded the alarm on the need to slow down, claiming that RSI “could outrun our ability to understand and control these systems, and so must be pursued very carefully, if at all.”
  • Behind public warnings, Anthropic has been building toward RSI all along. In 2023, a leaked investor deck revealed the company was privately pitching investors on a “next-gen algorithm for AI self-teaching.” By 2026, Anthropic openly claimed to be “delegating a growing share of AI development to AI systems themselves.” An independent report from METR provided evidence that Anthropic’s pursuit of RSI has dramatically sped up AI development – by a factor of nearly 2x.
  • Anthropic’s IPO filing shows the recursive loop is in full swing. Amidst warnings of catastrophic risk, the S-1 discloses that Claude is writing over 90% of its own code. The filing declares that “we may imminently meet our Responsible Scaling Policy’s definition of “dramatically accelerated progress attributable to AI.”
  • Despite public calls for regulation, Anthropic has voided safety commitments and weakened governance efforts. In 2026, Anthropic walked back its Responsible Scaling Policy, abandoning promises to pause unsafe development, and removing key tripwires for RSI. It refused to back California’s landmark AI safety bill until key provisions were stripped out. And experts have overwhelmingly criticized its recently leaked draft bill for weakening state efforts and failing to adequately protect against catastrophic risk.
  • Anthropic’s Head of Alignment Evan Hubingerwrote in 2026: “We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade.”

If you have any questions, please reach out to press@futureoflife.org.

This content was first published at futureoflife.org on November 18, 2026.

About the Future of Life Institute

The Future of Life Institute (FLI) is the world’s oldest and largest AI think tank, with a team of 35+ full-time staff operating across the US and Europe. FLI has been working to steer the development of transformative technologies towards benefitting life and away from extreme large-scale risks since its founding in 2014. Find out more about our mission or explore our work.

Our content

Related content

Other posts about 

If you enjoyed this content, you also might also be interested in:

Statement: The pro-human movement is winning

FLI's Head of U.S. Policy Michael Kleinman issues a statement on the U.S. Senate Midterm Election results.
5 November, 2026

Statement: We must pressure AI companies to immediately limit the use of recursive self improvement

FLI's Director of Policy Mark Brakel responds to "A Call for Control of Frontier AI Models", joint statement by heads of states.
28 September, 2026
Austrian Foreign Ministry/Michael Gruber

Government leaders call for treaty negotiations on autonomous weapons to begin in November

At a New York forum cosponsored by six governments, OHCHR, the ICRC, The Elders and FLI, ministers and UN officials called on states to begin negotiating a legally binding instrument on autonomous weapons at the CCW Review Conference in November.
28 September, 2026

Statement: World leaders at the UN call to rein in superintelligence race

Calls for pacing, coordination, and capacity-building are welcome, but the urgency of our shared situation requires practical solutions.
24 September, 2026
Our content

Sign up for the Future of Life Institute newsletter

Join 70,000+ others receiving periodic updates on our work and focus areas.
cloudmagnifiercrossarrow-up linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram