A top safety researcher at Anthropic has warned AI is advancing so quickly there is a greater than 10% chance it “could kill all humans” within the next decade.
Evan Hubinger said in a post on X that the risk from the models which currently exist was “low” but he was “worried” the technology may become able to improve itself soon to the point where it poses an existential risk to humanity.
It comes after the Financial Times reported Anthropic withheld its latest model from the UK’s AI Safety Institute, one of the leading bodies in the world for assessing AI risk.
The BBC has approached Anthropic for comment.
In his post, which has been viewed 9.6 million times, Hubinger added “we really do earnestly believe” AI poses a species-ending risk to humans.
“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 said.
Leading figures in the AI field have been raising the alarm about the safety threat the tech poses for years, with the heads of OpenAI, Google Deepmind and Anthropic saying as much in 2023.
But those warnings have become much more stark in recent weeks, as evidence emerges that firms may be struggling to control AI.
Over the summer, there were a string of incidents where AI agents – AI systems that are allowed to operate autonomously – carried out cyber-attacks.
OpenAI, Anthropic and Meta all disclosed hacks carried out by their AI tools.
And in September, OpenAI’s chief scientist Jakub Pachocki called for “extreme caution” over AI’s progress, warning more intervention may be needed to ensure “humans remain in control of the future”.
Major figures in the space have been calling for AI development to be slowed in recent months, including Anthropic bosses Dario Amodei and Jared Kaplan.
In an open letter signed by 1,300 staff members of AI firms, they called for the US government to “support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development”.
