xAI's Grok API adds custom content guidelines for NSFW, violence and profanity

xAI's API release notes say requests to the Chat Completions and Responses APIs can now carry a guidelines array that sets how Grok handles content in three categories: nsfw, violence and profanity. According to xAI, Grok follows the guidelines as it writes its response. The release notes point to a "Custom Moderation Controls" page in the docs for details.