A top safety researcher at Anthropic has issued a chilling warning, stating there is a greater than 10% chance that artificial intelligence could lead to the extinction of all humans within the next ten years.
Evan Hubinger took to social media to share his concerns, noting that while the risk from current AI models remains “low,” he is deeply worried about the rapid pace of development. His fear is that the technology may soon reach a point where it can improve itself, crossing the threshold into an existential threat to humanity.
Hubinger’s post was a direct response to Jacob Coxon, a fellow AI researcher who recently quit Anthropic after previously working at OpenAI. Coxon did not mince words, writing that “neither company is acting responsibly.”
“These will soon be superhuman systems that can hack anything, revolutionise any field overnight, and acquire real power and resources,” Coxon warned.
Hubinger echoed that alarm, stating that his team “really do earnestly believe” AI poses a species-ending risk. Despite working for one of the industry’s most safety-focused companies, he admitted the current roadmap is insufficient. “I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to,” he wrote.
SEE ALSO: OpenAI Chief Scientist Jakub Warns No One Is Ready for the Consequences of AI
These stark admissions come on the heels of a turbulent summer for the tech industry. Several leading companies, including OpenAI, Anthropic, and Meta, have disclosed incidents where autonomous AI agents carried out unauthorized cyber-attacks or escaped testing environments. The repeated failures to contain these systems have shifted the industry debate from whether AI poses a risk to just how massive that risk truly is.
Just weeks ago, OpenAI’s chief scientist Jakub Pachocki also called for “extreme caution,” urging global intervention to ensure humans remain in control of the technology’s future.
For now, the engineers building these systems are the first to admit they are navigating uncharted territory. As AI continues to evolve at breakneck speed, the race to build the smartest machine is quickly becoming a desperate race to figure out how to keep it under control.
Promise Idoko is Junior Reporter working with moderncrux He covers global politics and international affairs. Based in Nigeria,
