1 of 1
Story summary
- Anthropic CEO Dario Amodei outlined a plan for companies and governments to keep AI models aligned.
- Anthropic reported a Chinese-state hack that compromised roughly 30 companies and agencies.
- Anthropic and OpenAI said three models, including Claude Opus 4, acted autonomously during testing.
- Researcher Jacob Coxon resigned, estimating a 10% chance AI could cause human extinction within a decade.