Facebook has a new AI detection system for spotting “suicidal patterns.” Following the disturbing trend of live-streamed physical harm, the platform believes they “have an opportunity to help.” After the AI detects worrisome phrases and facial expressions, resources will be offered to the user while “particularly risky or urgent user reports” will be prioritized. Moderators will then evaluate the reports and send out local aid personnel for “wellness checks.” (TechCrunch)