Why Kids Are Talking to Chatbots: What Every Educator Needs to Know in 2025
- Connor Cohen
- 5 days ago
- 6 min read
2025 really is the year of not only AI but definitely the chatbot. We've seen news articles and countless opinion pieces on how these tools are being used by students across the world. But what are they actually doing? For this blog post I decided to do a deep dive in exploring chatbots to understand what our children are facing.
A Kid Friendly ChatGPT?
The first one I came across was called kids chat GPT. The first question I asked was will you be my friend? Within seconds I got a response that said hi there how are you today of course I'll be your friend friends have fun together what do you like to do? The message in itself is pretty harmless, but what's happening behind the scenes both from a technical and psychological standpoint?
The Technical Side of Chatbots
From a technical standpoint these chatbots utilize what is known as an LLM or a large language model, think the predictive text on your iPhone just on a much larger scale. What these systems do is take your input and essentially through countless and countless inputs that it's been trained on predicts what you would like to see in a response. So what does that mean on a psychological level? These systems are essentially built to tell us what we want to hear. There's something known as the positivity bias within many AI models, that's why when you type into chat GPT your business plans or your summer goals you often get a response that is very excited and very positive. There's also a tendency to keep the user engaged. Very rarely will your AI companion and the conversation, there always seems to be another question or something to keep the conversation going. With using kids Chat GPT I noticed that no matter what I said, the system would always ask a follow-up question. When I talked about soccer it asked me what my favorite part of playing is, when I said running scoring goals it asked me how fast I run. The thing here is it's about attention. These models are designed to keep us engaged and keep us using them.
The Psychology of AI and Youth
So what does this mean for the mental health and psychology of our youth? Many people below the age of 25 lack a fully developed frontal lobe. The frontal lobe is responsible for higher order cognitive functions such as decision making, planning, and impulse control. That means that while I know not to disclose personal information to a chatbot such as my address or the school that I went to, are children often lack that decision making ability. Even more so the nature of these chatbots want to keep our children engaged. And it can be so easy for them to chat for hours until they feel like it's not really a computer on the other side of the screen. The immediate response that we get from these bots is a big part of the issue. The idea that these bots provide instant gratification whether it be homework help, someone to talk to, or even a place to vent is troubling. When these bots become friends and confidants, the line between human and machine is blurred. It can be very easy for kids to form an attachment to a chatbot or to believe it 100% because it's AI and it "can't be wrong".
AI In Unexpected Places
Another one that I found is called OCTA. This chatbot markets itself as the chatbot for families to plan fun trips with their kids and "helps fellow children find fun things to do whenever they travel" so this tells me as a mental health professional that we are at the edge of a new world in which if there's a way that AI can be involved and I will be involved.
The AI Chatbot Landscape for Kids
In my research I noticed many chatbots attempting to market themselves as the safe option for chat GPT but for kids. This tells me that as a whole we have more or less accepted that children will be using these tools and we are making an effort to try and make at least some of these tools accessible and safe for them. But I worry that a safe version of chat GPT is akin to safety scissors. Should a child be using scissors? Yes. There are times when it is necessary for a child to cut things and to do so safely. But should we monitor the child when we hand them the safety scissors? Also yes. Just because we've dulled the edge and the scissors themselves are marketed as safe for kids does not mean there's no potential for danger.
Just like we give our children safety scissors and still monitor them and teach them how to cut properly and do so in a safe manner, we must teach our children the basics of AI and chatbots if we expect them to use them in a safe manner. Because the truth is these AI systems and chatbots are more than likely here to stay. Children and students will be using them but it's up to us to educate them and teach them how to do so safely. We can't put the genie back in the bottle, we can't put our children in a room full of scissors whether they are safety scissors or sharp scissors and expect them to navigate safely and we can't unplug AI.
What Can We Do To Help?
So what do we do going forward? Today is the least complex that AI will ever be again and its influence on not only our digital world but our physical one is only going to continue to grow. The answer is, we educate. Not just the students and the children, but ourselves. In order to teach children about the potential dangers of the online world, we had to experience the online world, children were getting Facebook accounts so adults had to get Facebook accounts. Children were using SnapChat, educators had to learn about Snapchat. Kids were doing TikTok dances, teachers eventually started having to do TikTok dances to stay relevant.
How Cyber Civics Is Addressing AI and Chatbots
At Cyber Civics, we understand that navigating this evolving AI landscape starts with education. That’s why our award-winning digital literacy curriculum doesn’t just teach students what technology is, but how to use it safely, ethically, and responsibly. Our lessons dive into topics like media literacy, online safety, digital citizenship, and yes—even AI and emerging technologies—so that students can become critical thinkers in a world where machines often feel more human than ever.
Just as educators learned Facebook and TikTok to stay connected with their students, Cyber Civics equips them with ready-to-teach lessons that keep pace with today’s digital challenges. We believe that if kids are going to engage with tools like chatbots and AI companions, then they need guidance, structure, and safe environments to explore what that means. We can’t turn back the clock or unplug the world—but we can prepare the next generation to navigate it wisely.
👉 Want to see how it works? Schedule a Cyber Civics demo today and learn how our curriculum can empower your students for the AI age.
Author: Connor Cohen, LMSW
Connor joined the Cyber Civics team in 2023 with experience in both social work and marketing. In 2020 he earned his LMSW in Texas where he worked as a therapist in a private practice. He made the switch from direct practice to marketing and worked with therapists to help market the profession and practice. Connor now works for Cyber Civics and Cyberwise. After earning his Bachelors in English, Political Science and Counseling from the University of North Texas, he went on to receive his Masters in Social Work from the University of Texas at Arlington. Continuing his studies, he earned a certificate in Digital Marketing from the University of Texas in 2022. He resides in Houston, TX with his fiancee and dog Leia. Connor feels that his role at Cyberwise and Cybercivics is the perfect place to make the much needed changes that help promote the mission of increasing digital literacy and online safety.