Anthropic CEO issues dire AI warning. Here's what he gets wrong.

Mashable • January 27, 2026 • business

Key Points:

Anthropic CEO Dario Amodei warns in a 38-page essay that super-intelligent AI capable of self-improvement could emerge within one to two years, posing risks such as human enslavement and mass destruction.
Amodei discusses potential threats including AI-driven bioterrorism, autonomous drone armies, and widespread job displacement, and calls for interventions ranging from industry self-regulation to constitutional amendments.
The essay anthropomorphizes AI by attributing motives and a sense of self-identity to current large language models (LLMs), a practice criticized as misleading since these models lack consciousness or genuine empathy.
Critics argue that AI doomsday predictions are often exaggerated sales pitches fueled by industry incentives, and that the focus should