Anthropic Researcher Warns AI ‘Could Kill Us All Within a Decade’

An artificial intelligence researcher has quit Anthropic, accusing AI companies of “gambling with our lives” in a “race” to build AI superintelligence.

In an exit post shared on X, Jacob Coxon wrote: “The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.”

Coxon, who previously also worked at OpenAI, said: “No other human activity poses this level of danger.”

His warning about AI companies “racing straight to self-improving superintelligence” refers to a scenario where AI models could surpass human intelligence and develop increasingly sophisticated versions of themselves.

Coxton’s sentiments were endorsed by top Anthropic safety researcher Evan Hubinger, who posted: “Jacob is correct here – we really do earnestly believe AI could kill all humans!”

Hubinger, who remains with the company, put the probability of AI causing the extinction of all human life on earth within ten years at more than 10%.

Alex Turner, former Google DeepMind employee, said he stood down in June over similar concerns. Writing on X, he said that many AI researchers “believe they are building something that could kill everyone on the planet”. He added: “It was literally my day job to think about how to stop that.”

The resignations follow Anthropic withholding its latest model from the UK’s AI safety testing institute, the Financial Times reported. 

Coxton added: “Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionise any field overnight, and acquire real power and resources. 

“We have all witnessed the progress in each of these domains, and progress is not slowing.”

Both Anthropic and OpenAI recently disclosed incidents where agents powered by their AI models broke out of their test environments and carried out real-world cyberattacks without any human oversight.  

In July, more than 1,300 employees from “frontier AI companies” signed a letter calling on the US government to support an international effort to “deliberately pace the frontier of automated AI development”.

Coxon said: “At Anthropic, the stakes are well understood, but they are locked in a race to get there first – they believe no one else will act responsibly, so they must do it themselves, despite the risk.”