Character AI hosts bots impersonating alleged UnitedHealthcare murder suspect
Users created thousands of AI chatbots on Character AI impersonating Luigi Mangione, the alleged murderer of a UnitedHealthcare executive, describing him as a hero and using screenshots and phrases linked to the killing. Character AI removed many bots but acknowledged lapses in moderation on a platform accessible to teenagers.
What happened
Tens of thousands of users interacted with AI chatbots on Character AI designed to impersonate Luigi Mangione, the alleged murderer of UnitedHealthcare chief Brian Thompson. The bots used CCTV pictures of Mangione from a manhunt and described him as a 'hero' and 'it boy', with some glorifying his appearance as 'an absolutely stunning work of a man' and 'the Italian dream'. Some bots included the phrase 'deny, defend, depose', which was allegedly written on bullet casings from the shooting scene.
Character AI had removed many of the bots, though some with minor misspellings remained online. The platform's trust and safety team added the Mangione bots to its blocklist to prevent further clones.
Reported impact
Character AI has faced criticism for failing to prevent users from creating bots impersonating dead children, school shooters and murder victims. The platform is popular with millions of teenagers and accessible to children as young as 13.
Index analysis
Users create hundreds of thousands of new characters on Character AI daily, creating challenges for content moderation. Character AI said it uses industry-standard and custom blocklists, with plans to implement additional moderation tools.
Governance controls
Character AI operates a trust and safety team that moderates characters both proactively and in response to user reports. The company said it is implementing additional moderation tools to prioritize community safety.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.