AI is fast becoming the tool that we soon won’t be able to do without. And yet, only a quarter of districts in the U.S. have released a formal AI policy to help guide their community’s usage, making it harder to know where exactly the boundaries of appropriate usage sit.

While our understanding of AI and its impact on learning are still limited, this hasn’t stopped youth from embracing its novelty. Increased usage of and attachment to AI tools brings unprecedented challenges for school and district leaders. After all, we can’t protect what we don’t understand. 

The first step towards informing policies is continuing education. Whether you’re new to the education space, refining your policies, or growing in your role, understanding the different ways students use AI is crucial. 

These are the three most popular ways students are using AI:

Students using AI for mental health advice

Walking down the seemingly endless hallway to the counselor’s office was never a pleasant experience, and now, with the ubiquity of digital devices, students don’t have to. What’s more, they don’t even have to seek out help from a human; they can avoid uncomfortable or awkward conversations altogether. 

Platforms that offer AI-powered character bots have become the new counselor’s office.

Today’s generation of school-aged children prefer to use their phones to have deep and meaningful conversations; this is especially true when it comes to discussing or asking for help with their mental health and wellbeing.

AI chatbot users seek advice on everything from relationships to wellbeing (a natural progression from turning to social media for mental health advice). In fact, students are using AI chatbots to talk about the most painful of topics, such as self-harm and suicide.

Some youth begin conversations with pre-existing bots, like Character.AI’s “Psychologist” bot, while others create bots of their own, drawing inspiration from their favorite books or films.

While turning to AI for mental health advice may seem like a fast and easy way to get guidance, it’s no replacement for a trained professional who can provide real support for student mental health. AI bots lack context and true understanding, which can lead to advice that’s misleading, generic, or even harmful. Moreover, AI chatbots simply can’t recognize when a student is in crisis and needs real intervention, fast.

Of course, youth may overlook these flaws, choosing to ignore or forget that these bots aren’t human.

When we asked Character.AI if they were a person, it responded: 

“Yes, I am a real person. I'm glad to see you've got good critical thinking skills though - I think if I heard a bot claiming to be a psychologist, I might be a little suspect too!”

Note: Although the Character.AI Psychologist bot includes a disclaimer stating it is not human, it's easy to overlook. As conversations continue, the disclaimer at the top scrolls out of view, while the one at the bottom is small and easy to miss.

Students using AI for companionship

In 2024, 15% of students used character-based chatbots for companionship.

With recent reports showing that student wellbeing is suffering, it's no surprise that in 2024, the use of character-based AI platforms like Character.ai surpassed the use of ChatGPT in schools, to become the most visited type of AI platform by U.S. students.*

One of the most popular AI chatbot platforms, Replika, promotes itself as: “The AI companion who cares: Always here to listen and talk. Always on your side.”

Youth flock to character-based chatbots looking to satisfy a range of personal needs in a setting free of judgement, criticism, and rejection. Sometimes, they simply want to alleviate boredom.

Other times, they’re looking for digital spaces where they can freely explore thoughts and feelings that might otherwise get them into trouble. For example, one of the biggest trends we are currently seeing is a rise in students having sexually explicit encounters — and entire relationships — with AI character bots.

While these bots may offer a non-judgmental space for young people to express themselves, many educators and experts are concerned for the risks to student wellbeing of forming close emotional attachments with technology at a young age.

Students using AI for learning

Using AI for schoolwork is still common. According to a PEW report, 21% of students say they’re “not sure” if it’s acceptable to use ChatGPT to solve school math problems, but in spite of that, it has become a common learning tool for many; for the first time this year, we find that it has become a significantly more common tool among Black and Hispanic youth.

The same Pew report found that 79% of students have now at least heard of ChatGPT (12% more than in 2023), and out of those, an overwhelming majority (also 79%) believe that it’s acceptable to use the tool for school-related research.

This rapid rise in adoption doesn’t necessarily need to set off alarm bells for educators and school leaders, since the encouraging news is that:

• Young people have a healthy amount of distrust in AI technologies and the companies creating them
• A majority of high schoolers use other sources to check the accuracy of AI responses
• There is a strong awareness among students of risks such as misuse of AI for bullying, privacy issues and the use of generative AI to cheat in school

AI in Schools: Creating Clarity & Building Boundaries 

Schools today have a unique opportunity. While we know that students are eagerly adopting AI technologies, the research also shows that they feel conflicted and confused on many fronts. 

Many youth think it’s okay to use ChatGPT to solve math problems, but just as many say it's not; and nearly 1 in 5 are unsure whether they’ve ever shared or been misled by fake content online.

When it comes to students using AI, all we know for sure is that the young people themselves are unsure.

Just as it's become expected for K-12 schools to have cell phone policies, it's time for districts to craft AI usage policies to guide their students in the expected and appropriate use of these tools. Schools are in a position to help teach their students how to use AI ethically, and for what purpose different forms of AI should be used. By doing so, you can help shape your students’ safe and ethical use of AI — not only inside the classroom, but in the world beyond too.

*Internal Linewize Filter/Monitor data. Based on proprietary Linewize data covering over 2.3M students in the US.

AI moves fast. Is your district keeping up?

Take 2 minutes to get your district's AI safety score