“AI is like a helper, like an enhancer,” says Dharunish Y. This high-school junior in Pennsylvania thinks AI-based chatbots, such as ChatGPT or Claude, can help people code, generate ideas and get their work done. Because of artificial intelligence (AI), he says, “I think that the number of innovations that happen in a specific time is just going to increase.” His positive outlook on the future of AI matches that of about one-third of all U.S. teens, according to a new survey.
The Pew Research Center asked 1,458 kids — all 13 to 17 years old — about how they use chatbots and how they view the likely impact of this type of AI.
Only 15 percent of teens — about one in seven — think AI will negatively impact their own lives. A larger share, 26 percent, think AI will negatively impact society. Abigail V. is one of them. A high-school sophomore, she attends the same school as Dharunish. “I worry that AI is going to be very negative,” she says. With plans to go into advertising, she’s concerned that AI will take away these types of jobs.
“AI can do exactly what [advertisers] can do, but just make it more artificial-looking,” she says. She also thinks AI is “kind of creepy.” That’s why she tries not to use it at all.
Dharunish, in contrast, is interested in computer science and has learned a lot about how AI works. He thinks this tech will definitely have a positive impact on his life: “I’m enthusiastic about it.”
These views and the results of the survey don’t surprise Linda Charmaraman. She directs the Youth, Media, & Wellbeing research lab at Wellesley College in Massachusetts. She had no role in the new survey. But she does regularly run workshops for teens about digital media, including AI. “Some of these teens,” she notes, “can teach the adults a thing or two about AI.”
Help with schoolwork
More than half of U.S. teens use AI, according to the survey — mainly to get information or help with schoolwork. Dharunish says he and his friends often use AI to walk them through solving difficult problems in math or chemistry. “I find it really helpful,” he says, “because I learn a new way to do a problem.” However, he adds, “I never let the AI do all the work for me.”
Not all students take that approach. Six in 10 of the surveyed teens say students at their school cheat with AI, using it either somewhat often or very often.
Abigail recalls witnessing such an incident in her marketing class. “I saw the person next to me asking what AI thought his opinion was of a video that we watched on advertising.” Instead of thinking about his own personal ideas, he had asked a robot. That’s taking things too far, Abigail says.
To avoid this type of situation, schools have to be careful to set boundaries on AI use, Charmaraman says. “The teacher wants their opinion — actually wants to know them better,” she says. It makes her sad that a student might feel that it’s better or easier to use a robot’s answer than to express their own thoughts.
Race and income level
Black and Hispanic teens use AI at higher rates than their white peers, the survey found. Only 8 percent of white teens say they get emotional support or advice from chatbots. Among Black teens, that rate more than doubles — to 21 percent.
Other studies, Charmaraman notes, have shown Black teens tend to spend more time online and on social media than white teens. “Households of color usually have more of a culture of acceptance of online worlds being a part of everyday life,” she says. In white households, she says, there’s often a stigma attached to spending too much time on tech.
Family income seems to matter, too, in chatbot use. Among households earning less than $30,000 per year, one in five teens say they do all or most of their schoolwork with AI’s help. That’s almost three times the rate seen in households where families earn $75,000 or more per year. Why? In low-income households and schools, Charmaraman says, parents and teachers are less likely to be available to help out. So AI might be starting to fill this gap.
The survey did not report any data about how students identifying as Asian American, Native American or LGBTQ+ used AI. Charmaraman says that in the future, she’d love to hear more about tech use among these groups.
Dharunish realizes AI has downsides. “I know there’s a lot of concern around how it’s used.” He hopes students keep up the hard work of learning. And he hopes people keep creating their own writing, art and music. “I don’t want AI to completely replace that,” he says. But, he adds, “I’d encourage people to still be open about what possibilities AI would bring.”