U.K. agency releases tools to test AI model safety

The U.K. Safety Institute has released Inspect, an open-source toolset designed to assess AI models' core knowledge and ability to reason. It aims to strengthen AI safety and can be extended and augmented via third-party packages written in Python.