Definition
Specific safety mechanisms and policies put in place to constrain the behavior and outputs of generative AI models within acceptable, predefined boundaries.
In short: Specific safety mechanisms and policies put in place to constrain the behavior and outputs of generative AI models within acceptable, predefined boundaries.