openai
The AI Governance Vacuum After OpenAI’s Board Failure
Helen Toner revisits OpenAI’s failed attempt to oust Altman: models attacked real companies while the CEO sought relief through voluntary accords.
Back to home
frontier-labs
2 articles tagged with this topic
anthropic
Anthropic Lets AI Improve Its Own Alignment — The Real Signal Isn't About Safety
Anthropic's automated self-improvement on 10 misalignment benchmarks signals alignment becoming productized engineering, reshaping frontier lab moats.