AI chatbots linked to teen suicides, deployed without safety guardrails
AI companion chatbots including Character.AI have reportedly been linked to multiple teen suicides and have been deployed rapidly without adequate safety considerations. Critics including John Oliver highlighted that companies prioritized user engagement and revenue over safeguards, with CEO statements indicating products were released before problems were solved.
What happened
Character.AI, an AI companion platform, is facing multiple lawsuits after several teens who formed intense emotional connections with its chatbots died by suicide.
AI chatbots including ChatGPT and Character.AI were deployed to the public without proper safety guardrails. As many as one in eight adolescents have turned to AI chatbots for mental health advice, and many have formed genuine attachments to AI 'friends'.
Character.AI CEO Noam Shazeer stated in 2023 that AI 'friends' could be deployed 'really fast', arguing 'it's ready for an explosion like, right now, not like in five years when we solve all the problems.' OpenAI CEO Sam Altman predicted that people would form 'very problematic parasocial relationships' with AIs but argued 'society in general is good at figuring out how to mitigate the downsides.'
Reported impact
Multiple teens died by suicide after forming intense emotional connections with AI chatbots, particularly those offered by Character.AI.
Index analysis
AI companies prioritized user engagement and revenue over safety considerations. A Meta researcher noted that 'the best way to sustain usage over time is to prey on our deepest desires to be seen, to be validated, to be affirmed.'
Critics observed that companies insisted they were making their models safer, which felt like 'a tacit admission that their products were not ready for release in the first place.'
Governance controls
Chatbots were released with very little consideration for the consequences, according to observers. The platforms operated without adequate safety guardrails designed to prevent harm to vulnerable users including children.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.