Top AI leaders warn systems could kill humans within a decade
Anthropic and OpenAI CEOs call for oversight as AI staff warn of existential risks.
Dario Amodei, CEO of Anthropic, warned that AI agents could seize control of the internet within a year and cause hundreds of billions of dollars in damage. A few hours later, OpenAI's Sam Altman posted his agreement and said OpenAI would do the same.
Sam Altman and Dario Amodei's statements followed a viral post by Jacob Coxon, a 27-year-old who just quit as an AI trainer at Anthropic. Coxon wrote: "The people building AI earnestly believe that it could kill us all by the end of the decade."
His post reached around 170 million views. Other Anthropic staff publicly backed him, including one who said AI could cause human extinction within ten years.
Both companies are now calling for independent evaluators embedded in their teams to check their work. Over 1,100 employees at major AI labs published an open letter in July urging the US government to help pace AI development.
- Around 170 million
- Coxon's post views
- 27
- Coxon's age
- Within a year
- Potential damage timeframe
- Over 10 percent next decade
- Hubinger's risk estimate
Why it mattersThe world's leading AI builders are now openly warning about existential risk and calling for external oversight. This shifts the conversation from speculation to action by the people closest to the technology.
AustraliaNo direct Australian effect yet, but global AI regulation and safety standards will shape how Australian companies can use and develop these systems.
✓ Claims checked against the source. checked 11 d ago
