# Who Is Jacob Coxon? Former Anthropic Researcher Quits and Warns AI Could 'Kill Us All' > Source: > Published: 2026-09-10 07:33:16+00:00 # Who Is Jacob Coxon? Former Anthropic Researcher Quits and Warns AI Could 'Kill Us All' ## Jacob Coxon's public warning went viral after he announced his decision to leave Anthropic A British artificial intelligence researcher who worked at both Anthropic and OpenAI has quit the AI industry with an extraordinary warning about where the technology could be heading. [Jacob Coxon](https://www.ibtimes.co.uk/ai-researcher-resigns-warns-superintelligence-threat-2030-1818808), 27, resigned from Anthropic on 9 September after roughly four months, arguing that leading AI companies are racing towards self-improving superintelligence despite potentially catastrophic risks. His warning that AI 'could kill us all by the end of the decade' subsequently went viral, prompting questions about who Coxon is, why he walked away and how he believes artificial intelligence could eventually threaten humanity. ## Who Is Jacob Coxon? Coxon is a British mathematician and AI researcher who studied at the University of Cambridge before entering the rapidly expanding artificial intelligence industry. He joined [OpenAI](https://www.ibtimes.co.uk/openai-chief-scientist-warns-ai-advancement-security-risks-1818185) around 2023, where his work included pre-training research and contributions connected to GPT-4o. Coxon moved to [rival AI company Anthropic](https://www.ibtimes.co.uk/anthropic-restricts-uk-access-claude-mythos-5-1-1818840) in July 2026, attracted partly by its reputation for taking AI safety seriously. However, he resigned months later and announced that he was leaving the industry entirely. He accused both companies of failing to act responsibly and claimed AI laboratories were 'racing straight to self-improving superintelligence and gambling with our lives.' ## How Does Jacob Coxon Think AI Could 'Kill Us All'? Speaking to Anderson Cooper, Coxon stressed that today's AI systems do not pose an extinction threat. His concern is what could happen if increasingly capable systems begin improving themselves without human involvement. He pointed to the rapid progression of AI in areas including mathematics, coding and autonomous tasks, arguing that systems which previously merely assisted humans are approaching the ability to perform some work independently. The most dangerous point, according to Coxon, would be [recursive self-improvement](https://www.ibtimes.co.uk/ai-governance-expert-warns-recursive-self-improvement-1818039). 'You could make it do the tasks that we're currently doing, and then you get what's called an intelligence explosion,' he explained. Under that scenario, an AI could conduct AI research itself, become progressively more capable and eventually reach intelligence far beyond human levels. Coxon argued that a sufficiently advanced autonomous system could potentially hack critical infrastructure or assist in creating catastrophic biological weapons. 'Right now, there's no risk of extinction,' he emphasised, instead warning that the rapid pace of development could make the situation considerably different within years. ## Anthropic Researcher Backs Coxon's AI Risk Warning Coxon's claims gained additional attention after current Anthropic researcher Evan Hubinger publicly backed the underlying concern. Hubinger said people working inside AI laboratories 'really do earnestly believe AI could kill all humans,' putting his own probability of that happening within the next decade at more than 10%. He similarly stressed that present models carry low extinction risk, with his concern focused on superintelligence emerging through recursive self-improvement. ## Why Did Jacob Coxon Leave Anthropic? Coxon said he initially intended simply to leave rather than launch a public campaign. 'I thought I can't be part of this anymore,' he told Cooper. He subsequently drafted an X thread with friends explaining his concerns. The post spread far beyond his expectations, accumulating tens of millions of views. His departure also carried a financial cost. Coxon reportedly left around two months before his Anthropic equity would have begun vesting, meaning he no longer has a financial interest tied to the company's valuation, although he retains OpenAI equity. Furthermore, Coxon believes competition between leading American AI companies and international rivals is pushing laboratories to move faster than they otherwise would. His warning ultimately concerns technology that does not yet exist. But his decision to abandon a career at two leading AI laboratories has given the wider debate over superintelligence and AI safety a prominent new voice. © Copyright IBTimes 2026. All rights reserved.