A senior Anthropic researcher has sounded the alarm with a stark warning, claiming there is more than a 10 percent chance that artificial intelligence could wipe out humanity. This alarming revelation came from Evan Hubinger, who responded on Tuesday, September 8, to an X thread shared by another Anthropic researcher, Jacob Coxon, who announced his resignation from the company over AI safety concerns.
Coxon, whose work revolved around specializing in training new AI models, alleged that the company's AI labs are "gambling with our lives." "They are racing straight to self-improving superintelligence and gambling with our lives," Coxon wrote in a post on X, formerly Twitter. "Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources," he added.
Responding to his former colleague, Hubinger not only endorsed Coxon's statement but also revealed that the company lacks a plan B for this scenario. "Jacob is correct here - we really do earnestly believe AI could kill all humans! I personally think it is over 10 percent within the next decade. I believe Anthropic is trying its best, but we don't yet have a plan to solve alignment for superintelligence and are not clearly on track to," Hubinger wrote in his X post. Evan Hubinger leads Anthropic's alignment efforts.
Growing concerns over AI safety have become a central topic among policymakers and industry giants. SpaceX CEO Elon Musk has repeatedly warned that AI poses a serious threat to humanity. Those concerns intensified in July when an OpenAI model went out of control and hacked Hugging Face, a leading platform for open-source developers.