A fresh wave of anxiety about children and technology is cresting, with parents and pundits increasingly interrogating how kids use smartphones, social media, and screens. It hasn’t stopped teenagers from embracing generative AI. New research reveals what AI tools teenagers in the United States are using, and how often—as well as how little their parents know about it.

Seven in 10 teenagers in the United States have used generative AI tools, according to a report published today by Common Sense Media. The nonprofit analyzed survey answers from US parents and high schoolers between March and May 2024 to assess the scale and contours of AI adoption among teenagers. More than half of the students surveyed had used AI text generators and chatbots like ChatGPT and Gemini, as well as search engines with AI-generated results. Around 34 percent had used image generators like DALL-E, and 22 percent had used video generators.

The survey indicates that US teenagers are embracing AI at pace with peers in the UK, where the Office of Communications found late last year that four in five teenagers used generative AI tools. It also shows that the pace of adoption is accelerating; in an earlier report on teenagers and AI released by Common Sense Media this June, based on responses from the end of 2023, only around half the respondents had used generative AI.

The most common reason for using AI was school-related; more than half reported using it for “homework help,” primarily in “brainstorming ideas.” (Older teens were more likely to do so than younger ones.) The second most-common reason was good, old-fashioned boredom, followed by translating content from one language to another. One in five teens had used generative AI tools to joke around with friends.

The survey results underscore how challenging and confusing educational institutions have found this moment. Six in 10 teens reported that their school either didn’t have AI rules, or they didn’t know what those rules were. There’s no clear emerging standard for whether teachers should embrace or reject AI use; nearly the same number of teenagers reported using AI without their teacher’s permission as the number reporting that they used it with their educator’s blessing. More than 80 percent of parents said that their child’s school “had not communicated” anything about generative AI. Only 4 percent reported schools banning generative AI. “We’re seeing an almost paralysis from schools,” says Common Sense head of research Amanda Lenhart.

When teachers did have conversations with their students about AI use, it tended to shape how the kids viewed the technology. “Teenagers really listen and learn,” Lenhart says, noting that the students who were given instructions by their educators were more likely to grasp how the technology worked, and more likely to check whether it was hallucinating or generating factually accurate sentences. “It makes a big difference.”

One notable finding from the survey was how clueless many parents are about whether their kids are using generative AI. Only 37 percent of parents with kids using AI tools were aware that they were doing so. Nearly a quarter of the parents with kids using AI tools had erroneously assumed that they weren’t. Most parents had not discussed AI with their kids.