If something you type triggers a safety filter, the conversation may be reviewed internally and stored for a period of time for moderation and safety purposes. In rare cases, authorities may become involved — but only if required by law or if there’s a serious, imminent threat that legally obligates a response.