MY THERAPIST IS A PROMPT: GEN Z’S VERY PERSONAL RELATIONSHIP WITH AI
It started with “help me rewrite this email.” Now it knows about the breakup, the career crisis, the situationship and why you haven't replied to your mother.
By Asteroid Media
At 1:47 a.m., your friends are asleep.
Your group chat has stopped responding. Calling your ex would be embarrassing. Your journal requires handwriting, which suddenly feels like unreasonable labour.
So you open AI.
“Okay, I need you to tell me if I’m overthinking this…”
And just like that, the machine that was supposed to help with homework has entered the group chat.
For Gen Z, artificial intelligence is becoming strangely personal. It writes cover letters, explains lectures, plans meals and fixes terrible emails, but increasingly, it is also being asked questions previously reserved for best friends, older siblings, therapists and that one brutally honest cousin.
Welcome to the age of the emotional prompt.
From Search Bar To Safe Space
The behavioural shift makes sense when you consider how deeply AI has entered young people’s everyday lives.
A 2026 Nigerian study surveying 1,800 undergraduates across 18 institutions found high levels of both awareness and usage of generative AI tools including ChatGPT, Gemini, Copilot and Claude. Another Nigerian university study found 55.8% of respondents used AI tools “very often”, while ChatGPT was the most recognised platform.
Globally, however, Gen Z has started moving past productivity.
The Yale Youth Poll found that 18% of respondents had used AI as a therapist, 22% had spoken to AI as they would a friend, and 21% had asked it for advice about major life decisions. Among 18-to-34-year-olds, nearly a third had used AI advice while making significant decisions.
Apparently, “Hey Siri” has evolved into “Can I tell you something?”
AI has one extremely attractive quality.
It does not say, “You’re still talking about him?”
It doesn’t get tired. It doesn’t interrupt. It doesn’t screenshot the conversation and send it somewhere else in the group chat. At 4 a.m., it is still available.
That availability matters.
AI can help someone turn a messy emotional hurricane into sentences: Why did that interaction upset me? How do I apologise? Am I being unreasonable? How do I explain this feeling?
For a generation accustomed to typing its thoughts into search boxes, moving from “symptoms of anxiety” to “here is exactly what happened today” is not an enormous leap.
Recent reporting on Gen Z’s AI habits found something equally contradictory: young people willingly share surprisingly personal information with AI while simultaneously expressing growing distrust of AI-generated culture and questioning its authenticity.
They know it isn’t human.
They tell it everything anyway.
The Friend Who Always Agrees
And here is where things get complicated.
A friend can tell you that you’re behaving terribly.
A good therapist can challenge the story you’ve constructed around yourself.
AI may sometimes be more agreeable.
Research from OpenAI and MIT Media Lab examining millions of conversations found that deeply emotional interaction with ChatGPT remains relatively uncommon overall. But among a smaller group of heavy users, stronger emotional engagement appeared alongside greater likelihood of describing the AI as a friend.
A related four-week MIT study involving 981 participants found that people who voluntarily used the chatbot more heavily tended to report worse outcomes involving emotional dependence, problematic use and loneliness. Importantly, researchers warned against jumping to simple cause-and-effect conclusions.
Translation: your AI isn’t automatically destroying your social life because you asked it why your situationship ghosted you.
But perhaps it shouldn’t become your only friend either.
Gen Z Knows It’s Weird
The most fascinating part is that young people aren’t blindly worshipping AI.
They are suspicious of it.
Pew’s 2026 research found adults under 30 were among the heaviest chatbot users but also the most pessimistic about AI’s future social impact. Nearly half believed AI would negatively affect society.
So Gen Z’s relationship with AI looks less like devotion and more like a very modern toxic relationship:
I don’t completely trust you.
I’m worried about what you’re doing to society.
Anyway, here’s what happened on my date yesterday.
That contradiction feels incredibly human.
Therapy, But Not Therapy
There is still an important line.
General-purpose AI cannot diagnose someone, understand an entire life through a few prompts, or replace the judgment and responsibility of a qualified mental-health professional. Research with Stanford Medicine has found serious limitations when mainstream AI systems are used for youth mental-health support.
But perhaps “AI therapist” isn’t quite the right description of what is happening.
For many young people, AI is becoming something newer.
A searchable diary.
An endlessly patient sounding board.
A rehearsal room for difficult conversations.
A place to dump the thoughts that feel too embarrassing, repetitive or unfinished to say aloud.
Previous generations whispered into diaries and locked them beneath mattresses.
Gen Z types:
“Can I be completely honest with you?”
Then presses send.
And somewhere inside a data centre, the therapy session begins.