In addition to TikTok and Instagram, teens are adding ChatGPT to the mix. About two-thirds of U.S. teens have tried an AI chatbot, and nearly one-third use one every day, according to Pew. Negative warnings about mental health are rubbish.
Pew Research Center on Tuesday released its latest survey on social media and internet use among teens, asking for the first time 13- to 17-year-olds how they interact with AI chatbots. Researchers found that 64 percent of young people self-reported being AI chatbot users, and 28 percent said they used AI at least once a day. 12% said they use AI several times a day, and 4% said they use AI “almost always.”
Unsurprisingly, OpenAI's ChatGPT is the go-to AI for teens, with 59 percent saying they've used it. Only 23 percent have used Google's Gemini, the second most popular AI chatbot, and Meta AI, Microsoft Copilot, Character.ai, and Anthropic's Claude are all used by fewer teens each year.
In an era when 97 percent of teens say they use the internet every day (according to a Pew study) and 40 percent describe themselves as “nearly always online,” it's no surprise that so many are embracing the hot new stuff in technology, especially as the pace at which AI companies push them into schools accelerates.
Microsoft, for example, is pushing CoPilot to schools in its home state of Washington, likely due to slow AI sales and the fact that only 14 percent of teens use CoPilot, according to Pew. OpenAI is similarly rolling out features for students, such as Learning Mode and ChatGPT for teachers, which launched last month, and which the company is making free until 2027 in an effort to branch out into academics before charging for the service.
The Trump administration has also pushed for greater use of AI in academic institutions, describing the technology as a way for the United States to remain competitive on the world stage.
Will someone please think about the children?
The Pew report focuses on usage metrics and does not include questions from American teens about how AI has impacted their personal lives or academic performance. I asked if they had any thoughts on the impact of AI on teenagers, but they did not respond.
However, there has been a lot of research recently done on this topic by other institutions, and these findings should be alarming when juxtaposed with data showing that two-thirds of teens are AI users.
In October, the Center for Democratic Technology (CDT) concluded that it had found a wealth of evidence suggesting that students were having troubling interactions with AI. According to the study, 42 percent of students have used AI for mental health support, companionship, or escape, and 19 percent said they or someone they know had developed a romantic relationship with a chatbot of their choice.
CDT found that most teachers have received little or no formal AI training and feel unprepared to deal with potential harm. Half of students also said that using AI in the classroom made them feel less connected to their teachers, suggesting that most are aware of these negative effects even as they continue to use AI.
A study by the Massachusetts Institute of Technology Media Lab also warned against the academic use of AI, reporting that students who wrote essays using ChatGPT over the summer had low knowledge retention rates. When connected to an EEG machine, the brains of students using AI were even shown to experience reduced stimulation. This suggests that bots are having a significant impact on how users think and their ability to learn while using technology.
AI chatbots are increasingly appearing in reports and complaints related to mental health crises, and not just for adults. Last year, a 14-year-old Character.ai user died by suicide, and his family sued the company, accusing its chatbot of playing a harmful role. In a separate lawsuit, the parents claim that ChatGPT further pushed their son into suicidal thoughts, which led to him taking his own life.
It's not a groundbreaking psychological study to conclude that teenagers are more sensitive than adults, nor is it a new finding that children are more susceptible to pressure from robots than older adults. But certainly, let's keep stuffing more AI in our kids' faces. What could go wrong?®
