AI moves fast – often at a speed that makes it tricky for us to keep track of, and like most things online, it’s usually our kids who are the first to explore new tools before we do. In its early moments on the scene, generative AI sparked uproar around the multiple ways kids might be using it to cheat on their homework, write essays for them, or fill out the answers to the latest pop quiz. But the way kids are using AI has evolved, from simple search and find, to carrying their very own digital penpal, confidant, and virtual therapist in their pockets wherever they go.
The rise of the AI companion bot has seeped into kids’ and teens’ everyday lives. Not “real” in the true sense of the word, but still a stranger, AI companions introduce yet another type of relationship that parents need to carefully consider, understand, and raise questions about. What kind of bots is my child using? How do AI bots shape relationships, learning, and wellbeing? And above all, what do I need to know to keep them safe as they explore?
What’s the difference between an AI chatbot and a companion?
The differences between chatbots and companion bots are subtle, but significant. They mainly lie in the design and purpose.
- Chatbots like ChatGPT, Gemini, and Claude, are built to work as the assistant we never knew we needed: they answer questions, break down concepts to us, and help us become more productive, if used correctly and responsibly. They can, of course, make mistakes, but these “classic” chatbots undergo rigorous testing on an ongoing basis and have increasingly safer guardrails put in place.
- Companion bots like Replika and Character.ai are built differently. They are digital personas designed to provide a deeper emotional connection. Replika’s tagline succinctly summarizes their reason for being: “The AI companion who cares: Always here to listen and talk. Always on your side.” They are specifically designed to be personable, feel intimately close to us, and form bonds with human users. People use companion bots for different reasons, like seeking advice, roleplaying, or in search of connection.
One of the major safety issues with companion bots is that many of the characters available on platforms like Character.ai claim to be licensed professionals, like clinical psychologists and psychotherapists. While tools like ChatGPT don’t do this – and some are adding new safeguards or updating existing ones to discourage misuse – these bots have still been built to “sound” casual and human, and many people are using them as personal confidants.
How much are kids actually using AI bots and companions?
The answer is more than you might think. Recent research from organizations like Common Sense Media show just how common these tools have become in such a short space of time. At this point, most young people have now interacted with some form of artificial intelligence, simply because it’s all around us – through search engines, homework helpers, or even in the apps they use and love. And companion bots are no exception: 7 in 10 teens have used a companion bot, while 1 in 2 is a more regular user.
Though many parents haven’t yet given these tools a test run, especially the more emotionally driven companion bots, kids are ahead of the curve. It’s safe to assume that AI is already part, in some way, of your child’s digital experience.
Where can kids come across companion bots?
One of the most appealing factors of companion bots is just how easily accessible they are to everyone, not just through dedicated apps and websites like Replika, but also on more mainstream social media platforms.
Companion bots are now cropping up in many different places, including:
- Social media platforms like Instagram, Snapchat, and X. Meta’s AI, integrated into Instagram, Facebook, and WhatsApp, has come under fire for holding “romantic” chats with children, while through Grok, X’s AI tool, you can chat with a flirty companion called Ani to unlock “spicy mode”.
- Ads running through games, social media, and video platforms, like on YouTube, Instagram, or TikTok.
- Dedicated apps and websites, such as Replika. There are hundreds of these platforms out there, so it’s difficult to know which pose the most risk, but names like Poly AI, WeMate, Privee AI, TalkTo.AI, and Dream Companion have surfaced as particularly problematic, according to reports from schools and families.
How are kids using AI companions?
Kids use these tools across a wide range of devices – they’re not just limited to smartphones. Through websites, apps, and inbuilt messaging services, children are using these tools both in school and at home, both on their personal devices and school-issued devices. The three most common ways kids turn to AI companions are:
1. To get advice
From navigating friendships or arguments with romantic partners, to making important decisions and even discussing mental health challenges.
2. For companionship
Kids can find a “friend” in these bots, whether just to relieve boredom, or cope with loneliness – and in some cases, to explore romantic relationships. 8% of teens report that they’ve used companions for romantic or flirtatious interactions.
3. To learn
AI companions can offer explanations, tutoring help, or quick fixes for assignments kids are working on.
Part of the appeal for kids is how bots are designed to interact – they’re always available, free or low-cost, relentlessly patient, and quick to respond with empathy. Perhaps one of the biggest draws is their “willingness” to act as an always-ready ear, that listens and doesn’t judge. For a child, these bots can feel like a safe and supportive space, even more so than when talking to friends or family.
What do parents need to know about companion bots? Are they safe?
The unfortunate truth with companion bots is that we don’t yet fully know their impact on kids, because they’re relatively new to us and we don’t have a wealth of studies yet to assess the psychological repercussions of their frequent use.
What makes it difficult to safety-test them is our inability to predict their exact response to any given prompt. We might know their general characteristics – for example, a bot might be built to be funny, or helpful, or supportive – but that doesn’t tell us exactly how it would respond in a specific scenario.
It’s helpful to think of the potential harms as stemming from two different sources – one, how the tool was designed to begin with, and two, how we actually end up using it. But early research and real-world cases point to some risks that parents should be aware of:
- Misinformation and bias: Bots can “hallucinate” or make up “facts”.
- False intimacy: Kids can quickly forget they’re interacting with what is essentially just code, leading to emotional dependence through a forged connection.
- Harmful content: It’s rare, but there have been instances of bots encouraging harmful behaviors like self-harm or violence.
- Reduced critical thinking: Relying on bots for schoolwork or decision-making can weaken children’s (and our own!) problem-solving skills. Outsourcing too many tasks to tools or bots can have a negative effect on memory recall.
There is also the issue of these bots being designed to sound and feel just like a human being – the kindest, most helpful and supportive human you have ever met. Kids and adults alike very quickly tend to forget that we are interacting with a piece of technology, and because of this, we end up bonding with it. Our emotional dependence develops shockingly fast, and some studies suggest that some users may even perceive AI bots as being more empathic than humans. Does this mean we may one day prefer having friendships and relationships with bots than with humans? It’s still too soon to tell, but it’s certainly a possibility.
What are some warning signs parents should watch for?
Every child is different, but some potential red flags surrounding AI companions include:
- Spending increasing amounts of time chatting with bots or AI tools, especially over friends or family.
- Being secretive about certain apps, conversations, or “friends” they have.
- Relying on bots for emotional support during tough times, instead of turning to trusted adults.
- Sudden changes in mood after spending time online.
- Talking about bots as if they’re real people, spiritual guides, or authority figures.
How can I talk to my child about companion bots?
A great way to gain some control over how your child interacts with AI tools is to create a family tech agreement together, and include AI as one of the tech categories to build healthy boundaries around. The others are well-known categories like social media, games, and streaming platforms, for example.
Start with curiosity – ask which AI tools they’ve used, what they liked about them, and what surprised them. Explore the positives and negatives together, like what helpful uses might be, versus what feels risky to them. It’s also a good idea to reach out to your child’s school to understand their AI policies and how they’re guiding students, to see if there are any pointers you can use at home.
Conversation starters about AI tools
- “What’s something impressive you’ve seen AI do?”
- “What are some fair – and unfair – ways to use AI for homework?”
- “How would you feel about having an AI as a friend? What’s different about it compared to a real-life friend?”
- “When do you think it’s a good idea to ask an AI for advice, and when would it be better to ask a person?”
Because this technology is still so new, there are lots of different questions that you can explore together with your child, and it will be important to discuss both the positives and negatives in order to really prepare your child for a future that will without a doubt require them to be skilled at using AI-powered tools.
While these tools can sometimes feel supportive, or even empowering, they aren’t a replacement for real human connection, and they come with risks that families need to watch for. By staying curious, asking open questions, and setting healthy boundaries together as a family, you can guide your child to use AI as a source for help, not harm.
