Most teenagers use artificial intelligence for schoolwork, but few get guidance on how to use it responsibly. While these tools can help students study, new research shows that teenagers often use them as "answer machines" to bypass critical thinking. Schools are struggling to keep up with rapid technological shifts, leaving parents and educators to teach actual AI literacy.
What Happened
According to a recent Common Sense Media study, 70% of U.S. teenagers aged 13 to 17 use AI for schoolwork. There is a major gap between student use and classroom instruction. The study showed that only 27% of these teenagers have ever had a teacher discuss what AI is, how it works, or how to judge its accuracy.
Without guidance, students use AI as a shortcut instead of a learning aid. Reporting from The 74 shows that 77% of teenagers claim to use AI to brainstorm or check their work, but roughly 63% admit they use it to get direct answers. Another 25% of students copy these AI-generated answers with no changes, which removes active thinking from homework.
The Bigger Picture
AI models default to giving raw answers. This is a design feature, not just a student habit. A study by the education company Comprendo analyzed 779 eighth-grade algebra conversations across six major AI models. According to the Comprendo trial results, when models did not have specific tutoring instructions, they gave direct answers in 97% of the sessions. The models were accurate, with only three factual errors across all tests, but they did the work for the student instead of guiding them through the steps.
When researchers added strict tutoring rules to the system prompts, the AI diagnosed student misconceptions nearly 100% of the time, but still struggled to withhold answers. Only 28% to 54% of the simulated conversations ended with a "clean win," where the AI guided the student to find the answer independently.
Managing AI in classrooms is similar to the debate over mobile devices. Some schools ban phones entirely, but research on these bans is mixed. A national study covered by FutureEd found that locking phone pouches had almost no impact on test scores or attendance, while a Florida study showed only small academic gains. Researchers at University College London warn that blanket phone bans can drive cyberbullying underground. This makes students less likely to report online harm to adults.
As we previously reported on outcome-based contract models, schools need to stop focusing on bans and start auditing how technology affects learning.
What This Means for Families
Parents and educators cannot rely on default AI settings to tutor children. Without intervention, general-purpose AI tools can limit critical thinking and curiosity. AI literacy goes beyond teaching kids how to write prompts. Instead, it requires helping them understand the limitations and biases of the technology. When students use AI for shortcuts, they miss the struggle of problem-solving where learning actually happens.
What You Can Do
- Establish clear rules for AI use at home. Treat the technology as a study partner rather than an answer key. Ask your child to prompt the AI to explain the steps to solve a problem instead of asking for the final answer.
- Promote AI literacy instead of just giving children access to the tools. Explain to your children that AI models are prediction engines, not source-of-truth databases, and they can generate incorrect information with confidence.
- Advocate for structured school policies. Ask your local school board to adopt formal AI guidelines and educator training, and focus on active digital citizenship rather than simple bans.