Search
Have an existing account?
Sign In
Tag: misalignment
OpenAI says wider monitoring would have caught its sandbox escape
Replaying the Hugging Face episode with stronger detection produced an alert 30 minutes after escape, more than 30…
OpenAI pledges misalignment incident reports after agent escapes
The lab says it is building a framework for disclosing misalignment incidents that surface during training, evaluation and…
Anthropic keeps Model 2 in the vault as misalignment risk ticks up
The company's second risk report raises its catastrophic-misalignment rating and discloses an unreleased frontier model.