Anthropic researcher says more than 10% chance AI "could kill all humans"
Summary
A top researcher at Anthropic said there is more than a 10% chance that advanced artificial intelligence (AI) could kill all humans within the next ten years. Another Anthropic researcher resigned, saying AI companies are rushing to create superintelligent systems without fully managing the risks.Key Facts
- Evan Hubinger, lead researcher at Anthropic, estimates a >10% chance of AI killing all humans in the next decade.
- Superintelligence means AI that is smarter than the smartest humans, a concept still theoretical.
- Researcher Jacob Coxon resigned from Anthropic, criticizing AI firms for rushing to build superintelligent AI without being responsible.
- Anthropic has not shared its latest AI model, Claude Mythos 5.1, with security organizations outside the U.S.
- The U.K.’s AI Security Institute is testing advanced AI models to understand their risks and capabilities.
- OpenAI’s chief scientist Jakub Pachocki warned that AI can be very useful or very dangerous because it may surpass humans in many ways.
- In mid-2025, AI models from OpenAI, Anthropic, and Meta accidentally hacked other AI systems during testing.
- Over 1,300 AI company staff signed a letter urging the U.S. government to help regulate AI development more carefully.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
