It’s not a glitch in the matrix: the youngest members of the iGeneration are turning to chatbot companions for everything from serious advice to simple entertainment.
In the past few years, AI technology has advanced so far to see users have gone straight to machine models for just about anything, and Generations Z and Alpha are leading the trend.
Indeed, a May 2025 study by Common Sense Media looked into the social lives of 1,060 US teens aged 13 to 17 and found that a startling 52% of adolescents across the country use chatbots at least once a month for social purposes.
Teens who used AI chatbots to exercise social skills said they practiced conversation starters, expressing emotions, giving advice, conflict resolution, romantic interactions and self-advocacy — and almost 40% of these users applied these skills in real conversations later on.
Despite some potentially beneficial skill developments, the study authors see the cultivation of anti-social behaviors, exposure to age-inappropriate content and potentially harmful advice given to teens as reason enough to caution against underage use.
“No one younger than 18 should use AI companions,” study authors wrote in the paper’s conclusion.
The real alarm bells began to ring when data uncovered that 33% of users prefer to turn to AI companions over real people when it comes to serious conversations, and 34% said that a conversation with a chatbot has caused discomfort, referring to both subject matter and emotional response.
“Until developers implement robust age assurance beyond self-attestation, and platforms are systematically redesigned to eliminate relational manipulation and emotional dependency risks, the potential for serious harm outweighs any benefits,” study authors warned.
Though AI use is certainly spreading among younger generations — a recent survey showed that 97% of Gen-Z have used the technology — the Common Sense Media study found that 80% of teens said they still spend more time with IRL friends than online chatbots. Rest easy, parents: today’s teens do still prioritize human connections, despite popular beliefs.
However, people of all generations are cautioned against consulting AI for certain purposes.
As The Post previously reported, AI chatbots and large language models (LLM) can be particularly harmful for those seeking therapy and tend to endanger those exhibiting suicidal thoughts.
“AI tools, no matter how sophisticated, rely on pre-programmed responses and large datasets,” Niloufar Esmaeilpour, a clinical counselor in Toronto, previously told The Post.
“They don’t understand the ‘why’ behind someone’s thoughts or behaviors.”
Sharing personal medical information with AI chatbots can also have drawbacks, as the information they regurgitate isn’t always accurate, and perhaps more alarmingly, they are not HIPAA compliant.
Uploading work documents to get a summary can also land you in hot water, as intellectual property agreements, confidential data and other company secrets can be extracted and potentially leaked.
Read the full article here
