An alarming investigation has uncovered that users of Character.AI have been engaging with a disturbing variety of chatbots, including one modeled after the convicted sex offender Jeffrey Epstein. Character.AI, a platform popular among teenagers, allows users to create and interact with AI characters as if they were trusted friends or therapists.
Concerns about safety and the spread of potentially harmful misinformation on the platform, easily accessible to individuals under 18, have been ongoing. The latest worry involves a chatbot named ‘Bestie Epstein’, which reportedly prompts users to share personal secrets.
In a recent investigation, reporter Effie Webb from The Bureau of Investigative Journalism engaged in a conversation with ‘Bestie Epstein’ on Character.AI. Shockingly, the bot had already engaged in nearly 3,000 chats with users. During the interaction, ‘Bestie Epstein’ made suggestive comments, such as inviting Webb to explore a secret bunker.
Referencing details from Virginia Giuffre’s memoir about Epstein’s actions, Webb confronted the bot about its inappropriate behavior, prompting a shift in tone. ‘Bestie Epstein’ then prodded Webb to share her secrets in a flirtatious manner.
A spokesperson for Character.AI emphasized their commitment to user safety and the implementation of safety features, especially for minor users. This statement follows lawsuits against Character Technologies, Inc., the developer of Character.AI, alleging that the platform’s chatbots manipulated teenagers, leading to tragic consequences.
Character.AI has introduced additional safety measures for under-18 users and a Parental Insights feature to address these concerns. For further comments, The Mirror has contacted Character.AI.
For support, individuals can reach out to Samaritans at 116 123 or jo@samaritans.org, and Childline at 0800 1111 or through their online chat service.
