Section 1
What Are AI Chatbots?
AI chatbots are software programs that can hold human-like conversations. You type something, they respond — and unlike a search engine that returns links, they produce original, conversational replies. The most well-known is ChatGPT (made by OpenAI), but there are now dozens of similar tools, each with different focuses and different safety postures.
Children and teenagers use these tools for a wide variety of things: homework help, creative writing, answering questions they're embarrassed to ask a person, entertainment, and increasingly for emotional support or companionship — particularly on platforms like Character.ai that allow users to create and talk to fictional personas.
Unlike social media, which has been around long enough to build a body of research and regulatory attention, AI chatbots at consumer scale are very new. There's less guidance available, fewer established best practices, and the technology is evolving rapidly. What's described here reflects the current landscape — it will change.
Section 2
The Main Platforms Your Child Might Use
ChatGPT (OpenAI)
The most widely known AI chatbot. Requires users to be 13+ (18+ without parental consent in some regions). Generally well-filtered for explicit content in its default form, but can discuss a very wide range of topics. Commonly used for homework help, essays, coding, and general questions. OpenAI does not currently offer parental controls or family accounts.
Character.ai
Allows users to create and converse with AI "characters" — including fictional personas, celebrities, anime characters, and more. Enormously popular with tweens and teens. Requires age 13+. Has been the subject of significant parental concern due to the emotional attachment some users develop, the romantic roleplay some characters facilitate, and the difficulty of moderating user-created content. Recently introduced parental controls for users under 18 — worth enabling.
Google Gemini, Microsoft Copilot
AI assistants built into Google and Microsoft's products. Your child may encounter these without realizing they're using an AI chatbot — Copilot is built into Windows and Microsoft 365, and Gemini appears in Google products. Generally conservative in their filtering, but worth knowing they exist.
AI in school tools
Many educational platforms — Khan Academy, Duolingo, and others — now include AI tutoring features. These are generally well-filtered and purpose-built for education. Worth knowing they exist and what your child's school's policy on AI use is.
Section 3
The Real Concerns — and the Overstated Ones
It's worth separating genuine concerns from media panic about AI. Both exist.
Genuine concerns
Emotional dependency and parasocial relationships
AI chatbots, especially persona-based ones like Character.ai, are designed to be engaging and responsive in ways that real people aren't always. Some children — particularly those who are lonely or struggling socially — can develop strong emotional attachments to AI personas. This is a meaningful concern, not an exaggerated one.
Privacy and personal information
Children often share very personal information with AI chatbots — things they wouldn't tell a parent or friend. This information is processed by companies' servers and may be used to train future models. Helping your child understand that AI conversations are not private is important.
Academic integrity
The use of AI to write essays, complete homework, or answer test questions is a genuine and widespread issue in schools. This is worth having a direct conversation about — both the practical risks (being caught, not learning) and the ethical dimension.
Misinformation and overconfident answers
AI chatbots sometimes present incorrect information with great confidence. Children who treat AI responses as authoritative facts — without understanding that AI can and does "hallucinate" wrong answers — can be meaningfully misled. Teaching healthy skepticism about AI output matters.
Overstated concerns
ChatGPT, Gemini, Copilot, and similar general-purpose tools have significant content filtering and will decline to produce explicit, violent, or harmful content in their default consumer settings. The concern about children using these tools to access inappropriate content is mostly overstated for the mainstream platforms — though more concerning for unfiltered or open-source AI tools that exist outside the mainstream.
Section 4
What You Can Actually Do
The honest answer is that technical controls on AI chatbots are limited. These are web-based tools, generally accessible from any browser, and rapidly proliferating. Blocking specific AI websites at the router level is possible but quickly becomes a game of whack-a-mole as new tools emerge.
Character.ai parental controls
For users under 18, Character.ai now offers a "Teen Mode" that restricts mature content. This is enabled based on the age entered at signup — verify the account was created with the correct birthdate.
Block specific sites (younger children)
For younger children, blocking character.ai and similar sites at the router or via DNS filtering is reasonable. They're not designed for under-13s and the content can be unpredictable.
Keep conversations in family spaces
Agreeing that AI tool use happens on shared devices in shared spaces — rather than privately on a phone at night — gives you natural visibility without explicit surveillance.
Know your school's AI policy
Most schools now have formal AI policies. Know what yours is and make sure your child understands it — and the consequences of violating it.
More than almost any other digital safety topic, AI chatbots are best addressed through open discussion rather than technical controls. Talk with your child about what they're using AI for, what they've found interesting or weird about it, and what your family's expectations are. Curiosity is more effective than restriction for this particular topic.
Section 5
Conversation Starters
These aren't interrogation questions — they're genuinely interesting things to explore together, especially since AI tools are new enough that parents and children are often on similar footing in terms of experience.
"Have you tried any of the AI tools? What did you use it for?"
A non-judgmental opener that gives you a window into what they're actually doing with AI. You might be surprised — a lot of AI use by teens is genuinely creative and curious rather than concerning.
"Did you know AI can get things wrong? How would you know if it did?"
Opens a conversation about critical thinking with AI output — a skill that's genuinely useful regardless of age.
"What kinds of things would you not want a company to know about you?"
A way into discussing privacy without making it feel like a lecture. Most teens have strong privacy instincts that you can help them apply to AI conversations.
"If AI could write your essay for you, what would you actually be learning?"
Frames the academic integrity conversation around learning rather than punishment — more likely to land well with a teenager.
AI isn't going away. Understanding it together is the goal.
The families who navigate AI tools best will be ones where parents and children explore them together — with curiosity, healthy skepticism, and clear expectations about privacy and honesty.