Anthropic researcher Jacob Coxon resigned after warning that frontier AI could pose an existential risk, including a more-than-10% chance of human extinction within the decade

Executive move ● Confirmed 87% confidence first seen 9 September 2026, 05:56

Anthropic researcher Jacob Coxon resigned and publicly criticized major AI labs, saying their safety efforts were inadequate and that systems they are building could lead to human extinction by the end of the decade. The resignation and his statements—particularly the cited “more than 10%” figure—prompted wide public and political debate about whether frontier labs should slow down development and strengthen oversight and evaluation.

AI 2040 Plan A AI 2040 slowdown plan AIUC AIUC-1 Andrew Yang Anthropic Astra model Barack Obama Bryan Cantrill Cabinet Office China Chris Lehane Claude Constellation Research Inc. Dario Amodei David Sacks DeepMind Demis Hassabis Denis Shilov Donald Trump Elon Musk Epoch AI Gallup Google Google DeepMind Holger Mueller Hugging Face JS Denain Jacob Coxon Jasmine Lambert Jensen Huang Jessica Ji Jonathan Kanter King Charles King Charles III MIT Technology Review Marc Benioff Mark Zuckerberg Medicines and Healthcare products Regulatory Agency (MHRA) Meta Microsoft Mike Johnson ModelTalk Mustafa Suleyman NVIDIA Nathan Lambert Noam Brown OpenAI Pew Research Center Philadelphia RSI (lossy self-improvement) Rajiv Dattani Robert Tappan Morris Rune Kvist Salesforce Sam Altman Siena University SpaceX Steve Eisman The New York Times U.S. Congress UK Government UK Joint Committee on Human Rights United States VicOne.Robot We Must Pace the Frontier White Circle White House Will Douglas Heaven Xi Jinping AI bill AI capability improvements AI data centers AI natives AI regulation AI safety and regulation AI safety regulation Agentforce All-In Summit All-In podcast Alphabet Amazon Americans Annenberg Public Policy Center Anthropic Claude Sonnet 5 Anthropic researcher Anthropic’s alignment lead Artificial Intelligence Act Artificial Intelligence Underwriting Company Associated Press Astra BBC Berkeley, California Bernie Sanders Bitcoin Bloomberg BloombergNEF CBS Sunday Morning Cabinet Office spokesperson Carnegie Mellon University Center for Security and Emerging Technology Claude Mythos Claude Sonnet Claude Sonnet 5 Codex Cornell Data Center Health Protection Pledge David Hubinger Decoder Democratic Party Disney Dreamforce Dreamforce Dumfries House Environmental Protection Agency Europe European Union Evan Hubinger Ezra Klein Facebook Faculty First Harmonic GPT-4 GPT-6 Germany Goldman Sachs Grace Huckins House of Representatives Humanist AI code of conduct Insight Media Group Ireland Jakub Pachocki Japan Jensen Huang says AI safety should be handled as an engineering issue by companies Lawrence Tallon Leonid Brezhnev Lionsgate METR METR Matt Burns Mythos 5 NVIDIA H100 Netflix Netherlands Niall Firth On Equity OpenAI model (unspecified) Paradox of the Worried West Paris Pew Research survey Politico Ribbit Capital Richard Nixon Ruben Bloom SAG-AFTRA SALT I Satya Nadella Scotland Sean O’Kane September 2026 Sherman Act SiliconANGLE SpaceXAI Steve Bannon The Ezra Klein Show The New Stack The Verge U.S. federal government UK Parliament UK government United Kingdom United Kingdom Government United States Congress United States Department of Justice United States House Judiciary Committee United States House of Representatives Vox WGAE Wall Street Washington University in St. Louis X agent-led cybersecurity incidents bioterrorism cyberattacks federal regulation government regulation independent evaluators independent monitoring persistent botnet sovereign cloud superintelligence third-party evaluation third-party evaluators xAI

Decision brief

What changed: Anthropic researcher Jacob Coxon resigned publicly after saying AI labs, including Anthropic, are not taking safety risks seriously enough. In the cited coverage, he warned there is more than a 10 percent chance advanced AI could kill all humans by the end of the decade and criticized continued work toward self-improving systems.
Why it matters: This raises reputational and governance pressure on frontier AI labs and any companies partnering with them, because a public resignation from a safety-linked researcher can intensify scrutiny of risk management, containment, and development pacing. For decision-makers, the immediate issue is not whether Coxon is correct, but whether stakeholder expectations around safety assurance, board oversight, and vendor accountability are rising faster than current practices.
Affected roles: CEO COO CTO CISO
Evidence: All three cited outlets report the same core facts: Coxon resigned from Anthropic and publicly warned that AI safety risks are being underestimated, including a stated greater-than-10 percent extinction risk by decade end. The Verge and TechCrunch are independent mainstream tech publications, and Trending Topics EU reports a consistent account, giving cross-outlet consistency on the event even though the claims themselves are primarily attributed to Coxon.
What remains uncertain: The coverage does not verify Coxon’s probability estimate, whether his views reflect broader internal consensus at Anthropic, or whether Anthropic will change policy or development pace in response. It is also unclear from the coverage what specific safety failures, internal escalation steps, or governance disagreements led to the resignation beyond the public statements reported.
Monitor next: Watch for any formal Anthropic response or policy change on safety governance, model containment, or release pacing following Coxon’s resignation.

Analytical support, not advice — assumptions and open questions stated above.

Source coverage

Anthropic researcher Jacob Coxon resigned after warning that frontier AI could pose an existential risk, including a more-than-10% chance of human extinction within the decade

Decision brief

Source coverage

Every AI story that matters, in your inbox by 8am.

Every AI story that matters,
in your inbox by 8am.