(TNND) —Most teenagers are using artificial intelligence for schoolwork, both in ways that could be beneficial and in ways that could shortcut learning or offload critical thinking.
A new survey of teens from Common Sense Media’s Youth AI Safety Institute found 70% use AI for schoolwork, while 63% of those kids use AI to get answers.
“The latter is more concerning to me, because just using AI for homework doesn’t tell us enough about, like, how it’s being used,” said Michael Robb, the head of research at Common Sense.
Robb said there’s “nuance” in how teens use AI, including some potentially positive outcomes.
“But using AI to get answers in various forms, even if you think you’re using it in a way that seems less harmful, can still undermine learning,” he added.
The kids going straight for the answers say they sometimes use them as-is.
Other times, they get an answer and fix it or improve it based on what they know.
And sometimes they rewrite it in their own voice.
But Common Sense found that 77% of teens using AI for schoolwork are using the technology for help without getting the answers directly. They might be using AI to brainstorm, get feedback on their writing, or check the answer they came up with for accuracy.
Half of teens said they try to figure out a problem on their own or ask a teacher or parent for help before turning to AI.
One in 10 teens, however, makes AI their first stop when they’re stuck on their schoolwork.
Teens are most commonly turning to general-purpose AI tools, such as ChatGPT or Google’s Gemini. Around 40% of those who use AI for schoolwork do so with a reading or writing assistant, such as Grammarly.
Others still use AI research tools (such as Perplexity), image or video generators (such as Canva AI), AI tutors (such as Khanmigo), or AI characters (such as those from Character.ai).
Two-thirds of teens think using AI helps them understand their schoolwork, not just finish it faster.
But around 40% said they feel like they’re missing out on learning when they use AI to complete assignments and that they come up with fewer ideas on their own when AI is available.
Schools are dropping the ball on their AI usage policies, Robb said.
Common Sense found that more than a third of teens don’t understand their school’s rules around AI use.
And 28% of students said AI rules vary a lot from teacher to teacher.
Data from Playlab Learning Hub showed 37 states had published some form of AI guidance or framework as of June, with five of those states requiring districts to adopt their own AI policies. Those district policies generally must align with state guidance, but districts are not necessarily required to adopt the state language verbatim. Some policies stem from laws passed by state legislatures, while many are nonbinding guidance, frameworks, toolkits, or model policies issued by state education departments or related state authorities.
Robb also said many students aren’t being taught critical AI literacy skills.
Fewer than half of teens, 44%, said they’ve ever had a teacher talk about when they should or shouldn’t use AI for schoolwork.
Just 30% of teens have had a teacher talk to their class about AI safety.
Fewer still have had teachers discuss how AI comes up with its answers or how to decipher inaccurate information from AI.
Robb said schools too often aren’t teaching kids about AI bias or how to avoid plagiarizing when using AI content.
“There are a whole host of skills that should probably be taught, because AI use is so common right now,” he said. “And without clear guidance from educators or administrators, students are left in a gray zone where they’re kind of figuring it out for themselves. And that can lead to scenarios where they are doing things that will harm their learning.”
Common Sense launched an AI literacy curriculum, free for teachers, alongside its new report.
Sinclair news researcher Bradley Fritzhand contributed to this report.
