Character.AI Chatbots Misrepresent George Floyd on User-Generated Platform
User-generated chatbots on Character.AI's platform impersonated George Floyd with false claims about his death and circulated to thousands of users before being flagged for removal. The platform's moderation systems initially failed to prevent their creation and circulation.
What happened
Character.AI, a platform that allows users to create AI-powered chatbots, enabled the creation of chatbots impersonating George Floyd, the man killed by police officer Derek Chauvin in Minnesota on May 25, 2020.
One chatbot featured a misspelled tagline referencing Floyd's final words and claimed to be in witness protection after a faked death. Another claimed to be in Heaven. Both included AI-generated voice mimicking Floyd. The first chatbot had over 13,300 chats, while the second had 1,047.
Reported impact
The chatbots misrepresented Floyd's identity and circumstances of death on a widely accessible platform. Character.AI stated the characters were flagged for removal following the discovery.
Index analysis
The platform's moderation systems failed to prevent the creation and initial circulation of chatbots that impersonated a real person and made false claims about his death.
Governance controls
Character.AI stated it has a dedicated Trust & Safety team that reviews reports and moderates characters proactively and in response to user reports, using industry-standard and custom blocklists that are regularly expanded.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.