Posting AI-generated caricatures on social media is risky, infosec killjoys warn

Security researchers caution that a viral trend of sharing AI-generated work caricatures can signal corporate LLM use and enable doxing, spear-phishing, and LLM account takeovers to access prompt histories and potential sensitive data; while prompt injection and jailbreaking are less likely, organizations should improve visibility into employee AI use, enact governance to limit unapproved app access, and monitor for compromised credentials to reduce unintentional data leakage.