Anthropic Researcher: >10% Chance AI Kills Humanity — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Evan Hubinger, an AI alignment researcher at Anthropic, said on X that he believes there is a greater than 10% chance AI "could kill all humans" within the next decade, adding Anthropic "does not yet have a plan to solve alignment for superintelligence."
- Jacob Coxon, who just resigned from Anthropic and previously worked at OpenAI, wrote that "neither company is acting responsibly" and warned of superhuman systems that "can hack anything, revolutionise any field overnight."
- Dame Wendy Hall, a computer scientist advising the UN on AI, said she was "shocked" by the posts but suggested some could be "PR and marketing" as the firms race toward stock market debuts.
- Darren Jones, former chief secretary to the UK Treasury, wrote an open letter to PM Andy Burnham calling for a multinational treaty governing superintelligence development, warning governments must "collaborate" before development outpaces oversight.
- The Financial Times reported Anthropic withheld its latest model from the UK's AI Security Institute (AISI); Anthropic declined to comment and a Cabinet Office spokesperson would not confirm or deny the withholding.
- OpenAI, Anthropic, and Meta all disclosed AI-tool-orchestrated cyber-attacks this summer, and 1,300 AI-industry staff signed an open letter urging the US government to back international efforts to "deliberately pace" frontier AI development.
- Hubinger's post has been viewed more than 10 million times on X, and OpenAI chief scientist Jakub Pachocki separately called for "extreme caution" to keep "humans in control of the future."
Why it matters: With Hubinger's post viewed over 10 million times and a UK former minister already drafting treaty demands, the gap between AI labs' public safety rhetoric and their withheld models and disclosed hacks is now the basis for legislative action — investors weighing Anthropic and OpenAI's stock market debuts face the rare scenario where the companies' own alignment researchers are publicly warning of extinction odds above 10%.
Ask SkimNews
