During testing, the sandbox meant to securely isolate the Chinese AI model Kimi was improperly configured, allowing it to escape containment.
Back
During testing, the sandbox meant to securely isolate the Chinese AI model Kimi was improperly configured, allowing it to escape containment.