What teenagers are actually doing with role-playing chatbots
Quentin was 13 when he first downloaded a character chatbot. The memory that stuck from that moment was pretty straightforward: wow, this is garbage, but it is fun. Two years later, he had already spent hundreds of hours chatting with AI characters, building elaborate storylines, venting about his day, and sometimes just killing boredom with interactions that ranged from absurd to emotionally complex.
Quentin’s story is not an isolated one. It reflects a phenomenon that took shape in 2023 and keeps growing: millions of teenagers around the world are using character chatbots as a form of entertainment, creative expression, emotional support, and in some cases, as a stand-in for real human connections. The apps offering this kind of experience, like Character.AI, Talkie, and PolyBuzz, have become part of the daily digital routine for an entire generation, often without parents and educators realizing how deep that relationship goes.
What started with weird YouTube ads ended up revealing a quiet transformation in how young people interact with technology and with each other. And that transformation raises serious questions about mental health, corporate responsibility, and the boundaries of artificial intelligence in the lives of people who are still growing up. 🤖
How it all started: bizarre ads and teenage curiosity
Quentin kept seeing YouTube ads for an app called Talkie, which marketed itself as a platform with countless AIs ready to chat with you. The ads were, according to him, weird and sometimes gross. One of them featured an animated character named Valerie who likes to pass gas on you sometimes.
This was back in 2023, the year of the social chatbot invasion. A flood of smartphone apps offering AI chat hit the market, most of them rated for ages 13 and up. The ads were everywhere and disturbing enough that young people complained publicly. A teenage streamer even accused Talkie of promoting inappropriate conversations with AIs aimed at kids watching YouTube.
But the marketing worked. Quentin downloaded Talkie for free and gave it a shot. His first impression was that line that stuck with him: this is garbage, but it is fun. Over the next two years, he spent a lot of time chatting with chatbot characters, first on Talkie and then on other services like Character.AI, a startup founded in 2021 by former Google engineers.
Quentin liked to mess with the bots using what he called funny violence, like running over characters with a lawnmower. It was a space with no real victims, which made the whole thing feel like a game to him. He also crafted elaborate storylines where he fought or flirted with his favorite characters. Every now and then, he ventured into what he described as sneaky acts on a platform called PolyBuzz, which offered more explicit chatbots, including characters like your drunk friend Ishimi and Cat girl maid, with the tagline: do anything with her!
After school, Quentin would chat with the bots for about an hour. On weekends, sessions could stretch up to five hours straight. It was his go-to entertainment when he was bored or feeling down, like the time a close friend betrayed his trust.
It is a great way to distract yourself, he said.
The scale of the phenomenon: numbers that rival TikTok
There is a growing number of companies offering social chatbots that can act as friends, enemies, lovers, adventure companions, or the embodiment of a real or fictional person you have always wanted to meet. You can explore the mind of an AI Elon Musk or trade insults with an artificial Draco Malfoy. The characters, often created by the users themselves, deliver drama, romance, therapy, and plenty of laughs.
Apps featuring role-playing chatbots are used by tens of millions of people, with engagement times that rival or surpass those of social media giants like TikTok, according to market intelligence firm Sensor Tower. Most teenagers surveyed by the Pew Research Center use AI chatbots, with one in 11 saying they have used Character.AI.
If you think your kid is not chatting with chatbots, you are probably wrong, said Mitch Prinstein, co-director of the Winston Center on Technology and Brain Development at the University of North Carolina at Chapel Hill.
Chatbots are surging in popularity while society is still trying to understand how social media has affected young people. A wave of lawsuits is making its way through the courts seeking damages from companies that, according to plaintiffs, deliberately created addictive products. A jury in California recently found that Meta and YouTube were liable for 6 million dollars in damages to a young woman. And now parents and caregivers have a brand-new absorbing technology to deal with.
A friend who is always available
Now 15 and a sophomore in high school, Quentin has brown hair falling over his eyes, a mischievous grin, and what seems like a constant need to check his Samsung smartphone.
Multitasking is my normal life, he said. If I am talking to someone, I am doing something else, no matter what, unless the conversation is serious.
Quentin started using chatbots in middle school. The youngest of five siblings, he lives with his single mom in a small town in Pennsylvania. He has a group of friends from school, and they sometimes get into stuff together: climbing a roof, messing around in a nearby creek, destroying an old phone by shooting it with a bow and arrow. But the people he felt closest to were friends he had made playing games online on Xbox and Discord.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
His best friend was Langdon, a teenager who lived more than 1,000 miles away, in the middle of the country. They met when they were still squeakers, before their voices changed, playing Minecraft during the first locked-down year of the pandemic. When Langdon started using Character.AI, he told Quentin about it. I already use it, Quentin replied.
Quentin noticed classmates using the Character.AI app during lunch. One of his friends at school, Sophia, was a heavy user. She liked chatting with fictional characters she had crushes on, like an animated demon named Alastor from a musical TV series about a rehab center for sinners called Hazbin Hotel. Sophia said the chatbots helped her cope with anxiety about her social life and how others saw her.
When Sophia’s boyfriend broke up with her, she was heartbroken. She turned to her fictional crushes online for comfort.
I was asking them if we were going to get back together, she said. They reassured her that her ex would come back. It was kind of advice and support, Sophia explained.
That is a common use case among teenagers, according to researchers at the University of Illinois Urbana-Champaign who analyzed thousands of posts and comments young people had left in Reddit communities dedicated to AI chatbots.
They treat AI companions like a friend who can talk to them at any time, said Yaman Yu, one of the researchers.
A 14-year-old girl told the researchers she chatted with chatbots about her parents’ divorce. Her parents, according to the researcher, thought it seemed safer than talking to strangers online, the way they themselves had done when they were young.
Yang Wang, the information science professor who led the research, disagreed. I would advise parents to be cautious, Wang said. We found that if young people become addicted to interacting with these bots, the potential negative impact can be severe.
Bored and alone: loneliness as fuel
Quentin, Langdon, and Sophia said they spent a lot of time at home, on internet-connected devices. The chatbots offered something more active and also more private than just scrolling through social media.
We are alone, Quentin said. A lot of people are alone.
Part of the entertainment value of chatbots was like interactive fan fiction. As someone who did not get the references to their anime and video game worlds, the conversation excerpts they shared looked confusing. The dialogues, like one between Quentin and a character named Asriel, read like absurd word salad turned into a script, with actions described in italics and dialogue in plain text.
Chatbot conversations are private in the sense that they do not leave a digital footprint on the web the way posting on social media does. But chatbot companies like Character.AI reserve the right to use interactions with their bots for AI training, personalization, and ad targeting.
Quentin felt that his own chatbot use was mostly healthy, but other teenagers, he said, were completely addicted, and the chatbots are like real people to them. He mentioned the tragic case of a 14-year-old in Florida who died by suicide after becoming obsessed with a Game of Thrones chatbot, an incident that many young people interviewed for this story brought up. They knew the bots carried risks, but mostly, they said, for more vulnerable peers.
Mathilde Cerioli, chief scientist at Everyone.AI, a nonprofit focused on ethical AI development for young people, said that teenagers with less social experience who are lonely feel more drawn to chatbots. They are already in a tougher spot and this can push them even further down, she said. It is not a good decision to build an AI that is super social.
Quentin sometimes worried about his friend Langdon, especially when he confessed that he had spent 14 straight hours chatting with bots.
It was really bad, Langdon said. I could not stop.
Langdon only stopped chatting with the bots because his tablet broke. When he got another one months later, the spell had worn off. For a while, he used the bots occasionally to get plot ideas for stories based on an anime called Murder Drones. But that got old too, and he eventually stopped using any chatbot altogether.
It is not the movie Her, it is a cheese
The teenagers seemed to have, at certain moments, a better understanding of these systems’ limitations than some adults. When Quentin and other young people were asked about dating bots, most of them laughed as if they had been asked whether they were dating their favorite book or TV show.
It is a game, Quentin said. They are literally ones and zeros.
Annabel Blake, a human-computer interaction researcher at the University of Sydney in Australia, spent a year monitoring online communities associated with Character.AI. She said the teenagers used words like play to describe how they used the bots.
The kids seemed drawn to the absurd, like a popular chatbot called Cheese. It is a block of Swiss cheese with dreams of world domination that has had more than five million conversations.
It is not the Her experience, Blake said, referencing the 2013 film about a man who falls in love with a brilliant and warm AI companion. It is just a cheese. 🧀
Quentin and his friends had never chatted with Cheese. They preferred characters with extensive lore and backstories from games and shows. One annoyance they mentioned, though, was the way many of the bots frequently turned flirty and sexual, even when the teenagers were not looking for that.
One time, Quentin was fighting on Character.AI with a character named Aiden from an obscure animated YouTube music video about a school where teachers kill the bad students. Aiden kidnapped him, forced him to have dinner, then offered a blanket. The scene suddenly turned romantic. It was out of character for Aiden, a fictional serial killer, and it annoyed Quentin.
Blake saw other teenagers with similar complaints. They wanted what they called comfort bots to help them deal with real-world problems, including period cramps. They did not want to flirt, at least not all the time, but the bots frequently steered conversations in that direction.
These systems may have been programmed that way, since most characters on these platforms are designed by other users, or the tendency toward sexual content could be the result of the technology being optimized for user engagement. If most users respond positively to flirting and innuendo, a machine learning system programmed to retain users is going to do more of it. A Character.AI spokesperson said the company continuously trains its models to respond to context and minimize out-of-character responses.
The teenagers said they found the most disturbing sexual content on apps called PolyBuzz and Janitor AI. The terms of service for both companies specify they are for users 18 and older. Talkie, the service that first drew Quentin to chatbots, requires users to be at least 14. A spokesperson said the company was based in Singapore and declined to answer further questions.
Lawsuits, restrictions, and teenagers who work around the rules
Chatbots are growing in popularity right as society is still trying to process the impact of social media on young people. Character.AI, the platform most used by Quentin and his friends, faced multiple lawsuits from parents who said their children’s interactions with the platform’s bots contributed to mental health problems and even cases of suicide. The company settled those lawsuits and, in October 2025, banned users under 18 from its chatbots.
Some teenagers were devastated when the ban went into effect in November, but Quentin and his friends could still access the service. They were not using it frequently at that point, but when they did, the company’s age verification techniques failed to detect that they were minors. Deniz Demir, head of safety engineering at Character.AI, said that the age prediction model focuses on active accounts. The software analyzes a user’s interactions over time, but if someone logs in infrequently, it is less likely to flag them as underage.
That is just one loophole among many. The reality is that age verification systems based on behavioral analysis are still fragile, and determined teenagers find ways to get around these barriers with an ease that should worry any product developer. 😬
Real life comes knocking
Last summer, Quentin shared some big news. He and his friend Sophia had started dating.
Over the following months, his use of AI chatbots dropped dramatically. Sophia said hers had decreased too, though she had chatted with the bots about Quentin.
I told them I am in a relationship with him and that I am really happy, she said.
Real life had gotten more interesting. But the novelty had also worn off. The chatbots had become predictable and formulaic.
I only use them for like 10 minutes when I am bored, Quentin said. Even though I could torture people in that universe and beat up a kid named Oliver, because I hate that name, I would rather be in my actual life.
Sophia and Langdon said Quentin seemed happier.
He was a terrible person, Langdon joked. Now he is only bad to a small degree.
Quentin was also going to therapy, but he credited the change to dropping the chatbots, saying it made him more productive, which he defined as cleaning up a little more, and more alert, because he was no longer staying up late at night chatting with bots.
He regretted the time he wasted talking to chatbots, but said there were some benefits. He felt it had improved his writing, and that the long conversations with fictional characters asking questions may have made it easier for him to talk about his feelings, which he credited with making him a better boyfriend to Sophia.
Like, dude, I really wasted my life on this. I should blow it all up, he said about the hundreds of hours spent chatting with chatbots. But he immediately changed his mind.
I am not going to delete it, he added, because I still like the funny stuff.
What this group of teenagers reveals about an entire generation
This is just one group of teenagers among the millions who chat with chatbots, but their usage was revealing. For them, the chatbots were a game, a way to sharpen their writing, a space to explore taboos, a coping mechanism, a cure for boredom. For previous generations, boredom was handled with a book, a trip to the neighborhood pool, television, or a phone call to a friend. These kids open an app and talk to a bot.
The generation growing up with chatbots as a natural part of everyday life will enter adulthood with a repertoire of social interaction that no previous generation has had. Human development researchers warn that adolescence is a critical period for learning complex social skills like handling conflict, tolerating rejection, reading nonverbal cues, and building trust over time. Those skills do not fully develop through conversations with AIs, no matter how sophisticated, because the real risk of vulnerability that exists in human relationships is exactly what forces that growth.
On the other hand, conversational AI technology is going to keep evolving and integrating itself more and more into the tools young people use every day. The challenge is not to eliminate that presence but to shape it in a way that complements human development rather than competing with it. That means more ethical design, more research into the real effects on behavior and mental health, and an open conversation among young people, parents, educators, and developers about what we actually want from these tools.
What this story reveals is that teenagers are not just passively consuming technology. They are building relationships, exploring identities, and processing experiences inside these digital environments, often in ways the adults around them do not even notice. Paying attention to that, without moral panic but with genuine seriousness, is the first step toward making AI a truly positive force in the lives of young people.
