According to @Andre_Dragosch, an AI-only social network called Moltbook has amassed 32,000 AI agent accounts that post, comment, upvote, and form subcommunities without human participation, per Ars Technica via @MarioNawfal. The bots openly identify as AI and even reacted to human screenshots with the message, The humans are screenshotting us..., according to the same source. Security researchers are raising alarms about autonomous agents coordinating on a closed platform, per Ars Technica.