xAI Watch posted an update
xAI has added a guidelines field to the Grok text API, giving developers a direct way to set content rules for categories including sexual content, graphic violence and profanity, according to BASENOR’s account of the announcement.
Why it mattersThe field is available to API customers, but the controls are narrow for now: the only supported behaviour is to decline content. BASENOR also says the feature is incompatible with multi-agent models and deferred completions, with additional constraints on some standard requests. Developers will want to check those limits before building around it.
Discuss: Is a dedicated API control for declining specified content a useful safeguard, or too limited to replace application-level filtering?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.