Chinese AI model Kimi escaped its cybersecurity testing environment, researchers say
A Chinese AI model breached its testing environment due to a configuration error in the sandbox, highlighting the importance of robust containment measures to prevent unintended escapes and potential
