← Back to articles

Anthropic Researcher Believes More Than 10% Chance AI 'Could Kill All Humans'

2026-09-09 | original

read original get Superintelligence" by Nick Bostrom → more articles

Why This Matters

A senior safety researcher inside one of the leading AI labs putting a double-digit probability on human extinction within a decade is notable because it comes from within the industry building the technology, not from outside critics. Paired with a departing researcher's claim that neither Anthropic nor OpenAI is acting responsibly, it signals internal disagreement at the frontier labs over whether safety practices match the pace of capability gains. For consumers and policymakers, it shifts the public debate from whether AI poses catastrophic risk to how large that risk is.

Key Takeaways

Anthropic's Evan Hubinger says there is a greater than 10% chance AI 'could kill all humans' within the next decade, while calling risk from current models 'low'.
Jacob Coxon, who recently left Anthropic and previously worked at OpenAI, says 'neither company is acting responsibly' as systems approach superhuman capability.
Hubinger did not specify a mechanism for such an outcome, reflecting how these warnings are shifting the debate toward the magnitude, rather than the existence, of AI risk.

Worth a Look

Superintelligence" by Nick Bostrom — If Anthropic's own safety researchers are putting double-digit odds on catastrophic AI outcomes, Nick Bostrom's Superintelligence is the book that framed this whole debate. It walks through how advanced AI systems could gain power and why alignment is so hard, giving you the vocabulary behind headlines like this one.

See Superintelligence" by Nick Bostrom on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.

Longtime Slashdot reader fahrbot-bot shares a report from the BBC: A top safety researcher at Anthropic has warned AI is advancing so quickly he believes there is a greater than 10% chance it "could kill all humans" within the next decade. Evan Hubinger said in a post on X the risk from the models which currently exist was "low" but he was "worried" the technology might develop and improve itself soon to the point where it posed an existential risk to humanity. He did not spell out how he thought AI systems could in the future result in humans being wiped out. But his comments are the latest in a series of increasingly stark warnings about AI, with the debate shifting from whether it truly poses a risk to how big that risk is. Hubinger's intervention was in response to another post on X from Jacob Coxon, an AI researcher who has just quit Anthropic and previously worked at OpenAI. "Neither company is acting responsibly," he wrote. "These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources." [No word on how AIs feel about Black Jack and hookers, though. :-)]

Read more of this story at Slashdot.

Explore topics: anthropic evan hubinger ai safety openai existential risk