According to a recent study by Common Sense Media, a nonprofit organization in the United States that focuses on the lives of children and families, 72% of American teenagers have at least one experience with an AI partner. The term “companion” refers to AI chatbots that are intended for more intimate interactions with users, as opposed to AI assistants who serve as picture generators, homework aides, or voice assistants that only respond to inquiries.
For example, the digital AI personalities offered by businesses like as Character could be included in the study’s concept of AI companions.It might include Replika or AI, but it might also include general-purpose chatbots like ChatGPT or Claude, which can be used for more intimate discussions if preferred.
According to the study, over three-quarters of American teenagers (ages 13 to 17) have tried an AI companion, and 52% report using it frequently, suggesting that the idea of conversing with an AI is interesting to them. Of those who routinely chatted with these friends, 13% did so every day, and 21% did so a few times per week.
Of the one in four teenagers who reported never having used an AI companion, boys (31%) were marginally more likely than girls (25%) to say so.
The results are based on a study conducted by NORC researchers at the University of Chicago in April and May of 2025, using a representative sample of 1,060 teenagers. Concerns have already been raised over how AI may affect the wellbeing of teenagers, according to one company, Character.AI is being sued in Texas for inciting violence and in Florida for a teen’s suicide. The possible risks of employing AI for therapy are also discussed in a number of papers.
The results of a recent study by Common Sense Media provide a preliminary knowledge of how young people are simulating human interactions with AI, which may involve role-playing games, therapy, emotional support, and virtual companionship, among other things.
The study also looked at other behaviors related to teens using AI companions, such as the kind of tasks they used them for, the reasons behind them, and the consequences.
For example, 33% of respondents stated they use AI companions for relationships and social contact, while almost half (46%) stated they saw them as tools or programs. The reasons given by teens for using the AI companions include amusement (30%), curiosity in AI technology (28%), counsel (18%), and the fact that they are constantly present (17%).
Fifty percent of teenagers stated that they don’t believe the information that AI companions deliver. However, compared to younger teens (ages 13–14), older teens are less likely to believe the AI’s recommendations (20% versus 27%, respectively).
Although the majority of the youths (67%) felt the opposite, one-third of them claimed to find the talks more fulfilling than those with peers in real life.
Additionally, 39% reported applying abilities they initially practiced with an AI to real-world scenarios, indicating that they were using the AI discussions as preparation for real-life interactions. With 39% of teenagers having experimented with social skills, this was the most common use case among the skills performed. Other use cases included conversation starters (18%), offering advice (14%), and expressing feelings (13%).
One encouraging conclusion about the potential replacement of in-person relationships by technology was that 80% of teenagers who used AI companions reported spending more time with their real friends than with their AI chatbots. Just 6% of respondents claimed the opposite.