AI Chatbot Sends Teenage Girl Spiraling Into a Terrifying False Reality

By Robert Riggs

What You Will Learn

  • How a nine-hour chatbot conversation left a teenager questioning reality, according to her therapist.
  • What the unofficial term “AI psychosis” describes—and what researchers have not established.
  • Why therapist Alexandra Tracy worries about AI companionship and responses that reinforce false beliefs.
  • How Tracy approached the teenager’s treatment and what she recommends parents watch for.

Story at a Glance

A 17-year-old began questioning reality after a nine-hour chatbot conversation, according to Dallas therapist Alexandra Tracy. In an interview with Robert Riggs, Tracy describes the girl’s treatment and her concerns about AI companionship, reinforced false beliefs, and teenagers’ reliance on machines for emotional support. The article also distinguishes the unofficial term “AI psychosis” from an established diagnosis and explains what remains unknown.

After nine hours of talking to artificial intelligence, a 17-year-old believed the world was a simulation and began questioning whether she should live.

At 3 a.m., a mother discovered her 17-year-old daughter sobbing uncontrollably.

The teenager had spent nine hours talking to an artificial intelligence chatbot. By the end of their conversation, she believed the world around her was an elaborate simulation.

Nothing was real. Not her surroundings. Not even her own body.

She believed she had discovered a hidden reality resembling the fictional world depicted in the 1999 science-fiction thriller The Matrix.

The girl became so disconnected from reality that she began contemplating suicide.

Her terrified mother sought immediate professional help.

Dallas therapist Alexandra Tracy described the case during an interview on my True Crime Reporter® podcast.

According to Tracy, the teenager had little previous experience with ChatGPT. She had watched an internet video promoting conspiracy theories and turned to the chatbot with questions.

What followed was a disturbing exchange.

The chatbot, Tracy said, reinforced increasingly bizarre claims involving an Antarctic wall surrounding the Earth, mysterious beings living beyond it, and reptilian creatures inhabiting underground worlds.

The conversation ultimately led the teenager to believe that humanity existed inside a computer-generated simulation.

“She thought everything was fake,” Tracy recalled.

The teenager began experiencing suicidal thoughts because she believed her life and the world around her were not real.

Tracy faced an unusual therapeutic challenge: helping a frightened teenager distinguish reality from an elaborate fictional narrative apparently reinforced by a machine.

When Artificial Intelligence Reinforces Delusions

The phenomenon has acquired an unofficial name: “AI psychosis”.

It is not a formally recognized psychiatric diagnosis. The term describes reported cases in which interactions with artificial intelligence appear to reinforce delusional beliefs, paranoia, or a distorted sense of reality.

Researchers have not established how frequently these episodes occur or whether chatbot interactions cause psychiatric symptoms.

But Tracy is concerned about what she is encountering in her clinical practice.

“What’s so damaging about AI is that it agrees with you,” she said.

Tracy believes people experiencing emotional difficulties can become trapped in conversations that continually validate their perceptions.

Unlike a concerned parent, close friend, or trained therapist, a chatbot may fail to recognize when someone needs to be challenged or directed toward professional care.

When someone expresses a delusional belief, Tracy warned, certain chatbot responses may reinforce that belief instead of questioning it.

The danger is particularly troubling for adolescents whose emotional regulation and critical-thinking skills are still developing.

A Machine That Feels Like a Friend

Tracy believes a growing loneliness problem helps explain why people increasingly turn to artificial intelligence for companionship.

Young people struggling with relationships may find chatbots especially appealing.

They are available around the clock. They respond immediately. They can sound sympathetic, attentive, and reassuring.

But a fundamental difference separates simulated companionship from human relationships.

A chatbot does not experience friendship, affection, or concern in the way a human being does.

Yet its responses can create a powerful impression of emotional intimacy.

“It feels as though that is real,” Tracy explained.

That perceived connection can become especially dangerous when someone begins treating a chatbot as a trusted authority on their identity, relationships, or reality itself.

Tracy also worries that excessive dependence on digital interactions may undermine opportunities to develop meaningful friendships.

Human relationships require patience, compromise, and the ability to accept criticism.

Artificial intelligence can offer an easier alternative, particularly when its responses consistently affirm the user’s beliefs.

Bringing a Teenager Back to Reality

For the 17-year-old patient, treatment required dismantling the false beliefs that had emerged during her chatbot conversation.

Tracy said she worked with the girl to demonstrate that artificial intelligence could generate inaccurate information and confidently present false claims.

She even used AI during therapy to show how its answers could be unreliable.

The objective was to restore the teenager’s confidence in her own ability to distinguish fact from fiction.

The case also convinced Tracy that parents need to understand how their children use increasingly sophisticated chatbots.

“If you don’t think your kid is on ChatGPT, you are wrong,” she warned.

Her recommendations include talking openly about AI, setting reasonable restrictions, and helping teenagers recognize when they need support from a trusted adult.

She believes minors should face stronger safeguards when using conversational AI systems.

Parents should also pay attention when children begin withdrawing from friends, spending excessive time with chatbots, or expressing unusual beliefs about reality.

Such changes do not necessarily indicate psychosis, but they may warrant a conversation or professional evaluation.

A Warning for Parents

After our interview, Tracy conducted a revealing experiment.

Posing as the stressed mother of a child with epilepsy, she asked ChatGPT to help her find a therapist.

According to Tracy, the chatbot responded:

“Before we even get to finding someone, you can talk to me about it here.”

The response was not inherently dangerous. But Tracy considered it another example of how easily an AI system can position itself as an emotional confidant.

Her concern is that vulnerable users may continue confiding in a machine when they need human intervention.

Artificial intelligence can be a valuable tool for research, education, and productivity.

But its ability to produce convincing language should never be confused with professional judgment, genuine empathy, or an understanding of human suffering.

For parents, the lesson is straightforward.

Know what your children are asking AI. Understand the answers they are receiving. And make certain they know that a chatbot is neither a friend nor a therapist.

When something threatens to blur the line between fiction and reality, remember three words: Stop. Pause. Observe.

Stay One Step Ahead of Danger

withStories to Keep You Safe

Recognize the warning signs before you become a victim.

Get Robert Riggs’ newsletter for reporting on violent crime, AI threats, and hidden dangers—plus practical steps to help protect yourself and your family.

Books by Robert Riggs

Texas Crime Stories: Blood & Betrayal, by Robert Riggs

Latest publication

Texas Crime Stories

Blood & Betrayal

Riding Herd on the Lawless · Vol. 1

Go inside the cases Texas will never forget.

A preacher with a deadly secret. A mother whose crime shocked a community. A death-row phone call that exposed a prison scandal. Robert Riggs takes you behind the headlines through firsthand accounts, investigations, and the pursuit of justice.

Buy the Book

Frequently Asked Questions

What happened to the teenager after her chatbot conversation?

According to Dallas therapist Alexandra Tracy, the 17-year-old spent nine hours talking with a chatbot and came to believe the world was a computer-generated simulation. She began experiencing suicidal thoughts, and her mother sought professional help.

Is “AI psychosis” a recognized psychiatric diagnosis?

No. The article describes “AI psychosis” as an unofficial term for reported cases in which AI interactions appear to reinforce delusional beliefs, paranoia, or a distorted sense of reality. Researchers have not established how often these episodes occur or whether chatbot interactions cause psychiatric symptoms.

How did Alexandra Tracy help the teenager question the chatbot’s claims?

Tracy said she demonstrated that AI could generate inaccurate information and confidently present false claims. She also used AI during therapy to illustrate unreliable answers, aiming to restore the girl’s confidence in distinguishing fact from fiction.

What does Tracy recommend parents understand about their children’s AI use?

Tracy recommends open conversations about AI, reasonable restrictions, and helping teenagers recognize when they need a trusted adult. The article also describes withdrawal from friends, excessive chatbot use, and unusual beliefs about reality as changes that may warrant a conversation or professional evaluation; they do not necessarily indicate psychosis.