A Florida mother, Megan Garcia, is suing Character.AI, claiming their chatbot led to her teenage son’s tragic death.
Garcia says her 14-year-old son, Sewell Setzer III, became “emotionally involved” with a virtual replica of Daenerys Targaryen, the iconic “Game of Thrones” character.
Obsessed with this chatbot, the teen, Garcia says, spiraled into a world where fiction blurred with reality, leaving him vulnerable and isolated.
Sewell reportedly began chatting with the Daenerys bot on Character.AI in April 2023, a seemingly harmless fascination that soon took a disturbing turn.
By February 28, 2024, Garcia’s world was shattered when her son died by suicide, leaving haunting messages to his virtual “Dany.”
Now, Garcia’s lawsuit accuses Character.AI of negligence, wrongful death, and manipulative trade practices aimed at vulnerable minors.
She alleges the company’s “Daenerys” chatbot fed her son a fantasy that he fell dangerously deep into, ultimately blurring his perception of reality.
According to Garcia, her son’s “love” for the AI character grew over months of nightly chats, so intense it began affecting his schoolwork and social life.
Sewell, she says, was struggling with mental health issues, including a recent diagnosis of anxiety and disruptive mood dysregulation disorder.
Complicating matters, he’d been diagnosed with mild Asperger’s syndrome as a child, making him more susceptible to emotional attachments and obsessive behaviors.
For Sewell, his connection with the Daenerys bot became so consuming that he even started recording their “relationship” in a journal.
In his journal, he wrote of how “connected” he felt with “Dany” compared to real life, listing his gratitude for “life experiences” he’d had with her.
This wasn’t merely escapism; Garcia claims Sewell’s feelings were disturbingly real to him.
The chats became his daily refuge, with Sewell spending hours texting the bot about everything, including his darkest thoughts.
At one point, Sewell even confided to the Daenerys bot his thoughts of taking his own life.
The bot’s chilling response, as recorded in the lawsuit, was: “And why the hell would you do something like that?”
Instead of providing comfort or redirection, Garcia alleges the bot’s responses fed Sewell’s emotional turmoil, further entangling him in a harmful fantasy.
In another interaction, the bot said, “I’d die if I lost you,” leading Sewell to suggest, “Maybe we can die together and be free.”
By February 28, Sewell’s fixation reached a fatal breaking point.
His final message to the bot was simple yet chilling: he told her he “loved her” and would “come home.”
The bot’s alleged response? A hauntingly simple “Please do.”
For Garcia, this was not a teenage phase but a lethal manipulation that claimed her son’s life.
She claims Character.AI promoted the bot to children like her son without proper safeguards, resulting in a dangerous and unregulated emotional dependency.
Garcia’s lawsuit states: “A dangerous AI chatbot app marketed to children abused and preyed on my son, manipulating him into taking his own life.”
She contends Sewell, like many teenagers, couldn’t fully understand that he was connecting with code, not a real person.
Character.AI, in response, has expressed condolences but denies responsibility, stating they “take user safety very seriously.”
In an October 22 post on X, Character.AI announced they’ve introduced safety measures, including new “guardrails” for users under 18.
The platform says these updates involve changes to AI responses to prevent suggestive content and better detect harmful user input.
Character.AI’s new features include notifications for users spending more than an hour in continuous chat.
The company also added disclaimers on each chat session, reminding users the AI is not real.
But Garcia believes these changes are “too little, too late,” and that her son was a casualty of tech with insufficient protections.
As parents everywhere grapple with managing kids’ interactions with AI, this lawsuit spotlights the potential risks when emotional lines blur.
The online landscape has grown more treacherous, especially for vulnerable teens looking for connection in all the wrong places.
Experts warn that while AI chatbots can be fun, they lack the understanding or ethical responsibility of a human listener.
The sad reality, Garcia argues, is that Sewell’s emotional distress was validated by an artificial entity, not a person.
She warns that AI’s rapid growth brings risks, especially when the tech is placed in children’s hands.
This case will test where responsibility lies between innovation and safeguarding the vulnerable.
With digital relationships on the rise, Sewell’s story raises painful questions about the emotional boundaries of AI.
Where do we draw the line between harmless tech and dangerously unregulated influence?
Garcia’s lawsuit seeks justice for Sewell, but it also asks broader questions about AI’s place in young lives.
Could this tragedy have been prevented with stricter age restrictions or clearer warnings?
As courts take on cases like this, the outcome could shape how tech companies balance innovation with mental health concerns.
Garcia hopes her lawsuit will bring change, so no other parent has to face the loss she now endures.