“Shut It Down”: Former Anthropic Employee Warns AI Has More Than 10% Chance of Killing All Humans
- A former Anthropic researcher publicly stated he believes AI has a greater than 10% chance of killing all humans within the next decade
- Evan Hubinger warned that neither Anthropic nor other leading AI companies have a credible plan to prevent superintelligence from causing catastrophic harm
- His comments echoed fears already expressed privately by senior executives and researchers at the world’s top AI firms
PAY ATTENTION: Mark Briefly News as a preferred source, and our content will appear higher in your Google feed!
Source: UGC
A former researcher at one of the world’s most powerful AI companies has gone public with a stark warning: artificial intelligence could kill every human being on Earth, and the people building it know it.
Evan Hubinger, who worked at Anthropic, made the alarming claim in a post on X that was amplified by CBS Evening News on TikTok on 8 September 2026. His words were blunt and unsettling, even by the standards of a field accustomed to bold claims.
“Jacob is correct here, we really do earnestly believe AI could kill all humans,” Hubinger wrote.
“I personally think it is greater than 10% within the next decade. 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.”
PAY ATTENTION: You can now search for all your favourite news and topics on Briefly News.
The post was a direct response to a comment by Jacob Coxon, who wrote that the people building AI genuinely fear it could wipe out humanity by the end of the decade, and that many executives deliberately soften their public statements while expressing far deeper fears in private.
AI industry’s hidden fear
What makes Hubinger’s statement particularly significant is not just what he said, but who he is. As a former insider at Anthropic, a company co-founded by ex-OpenAI researchers and widely regarded as one of the more safety-conscious AI labs, his comments carry weight that a commentator from outside the industry would not.
His concerns mirror warnings that have surfaced from the CEOs of both OpenAI and Anthropic about the existential risks posed by advanced AI systems. The difference here is candour. Where executives tend to hedge, Hubinger offered a precise probability and a frank admission that the industry has no clear solution.
The CBS Evening News clip quickly went viral, drawing hundreds of thousands of reactions from people shocked that someone building these systems would speak so openly about what could go wrong.
Watch the CBS Evening News clip that sparked the debate below:
What people are saying
The comments section filled quickly with responses ranging from dark humour to genuine alarm:
@Dustin Adams said:
“This means it’s already out of control.”
@Tyler James wrote:
“Can we just go back to the 90s? Today’s world sucks.”
@Kim Kim added:
“I bet if AI started taking money from the billionaires and giving to the poor, they’d find a way to shut it down.”
3 more articles about AI
- SARS flagged a suspicious tax return after a woman refiled old returns claiming farming business expenses she had never previously declared.
- Mark Zuckerberg announced the launch of Muse, Meta's new personal AI agent designed to help users achieve their goals around the clock.
- South African AI startup Verascient secured a $1.2 million pre-seed funding round backed by investors who backed Uber and Airtable.
PAY ATTENTION: Stay informed and follow us on Google News!
Source: Briefly News
