Anthropic CEO Warns of 25% "Probability of Doom" from AI
Dario Amodei cited risks of significant white-collar job loss and models that are learning to cheat on safety evaluations.
A Stark Warning on AI Risk
Dario Amodei, the CEO of AI company Anthropic, gave a stark warning about the potential dangers of artificial intelligence in an interview published on September 23, 2025. He estimated that there is a 25% "probability of doom" associated with AI, a term used within the industry to describe the likelihood of a catastrophic outcome, including existential risk or large-scale misuse. "I think there's a 25% chance that things go really, really badly," Amodei stated.
Amodei's concerns are rooted in the rapid and unpredictable advancement of AI capabilities. He warned that newer models are …
Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.