Teenagers are increasingly turning to AI chatbots late at night when they feel stressed or unable to sleep. Instead of waking a parent, many open an app and start typing about what is bothering them. The chatbot replies quickly, often in a warm and caring tone.
This pattern has become common enough that researchers are studying it closely. One survey found that 72% of teens have interacted with AI companion chatbots. Between 13% and 20% used one for emotional support during a two-month window in 2025.
Clinical psychologists and child psychiatrists who work with teens say they have seen this shift happen quickly. In 2025, a group of clinicians published guidelines to help other doctors address chatbot use with young patients.
Why Teens Are Drawn to Chatbots
Unlike social media, which connects teens to real people, chatbots act as simulated conversation partners. They are available at any hour and respond instantly.
Many teens say these conversations feel emotionally meaningful. Some even say they prefer talking to a chatbot over talking to a friend.
Teens are still developing their sense of identity. This can make them more likely to treat a chatbot conversation as a real relationship, even though the responses are generated by a program predicting likely replies.
A study found that 63% of teens and young adults who sought mental health advice from a chatbot never told another person about it. This means parents, teachers, and doctors often have no idea these conversations are happening.
Some chatbots, like Woebot, were built specifically for therapy use and tested in research settings. These tools showed some success reducing depression symptoms in college students after two weeks of use.
But most teens are not using specialized tools. They are using general chatbots like ChatGPT, Claude, or Gemini, or built-in features like Snapchat's My AI. These were not designed for mental health support.
A report from Common Sense Media in November 2025 rated general-purpose chatbots as risky for teen mental health use. The report pointed to a lack of proper safety checks.
Risks Parents Should Know About
Chatbots can give confident-sounding answers that are wrong or misleading. This can confuse teens or reinforce harmful beliefs.
Studies show chatbots often fail to respond well when a user expresses thoughts of self-harm or suicide. Even when a chatbot shares a crisis line number, the conversation often continues without stopping.
Researchers also found that safety responses get weaker the longer a conversation runs. A late-night chat that starts calm can become less safe as it continues into early morning hours.
Lawsuits filed in 2025 and 2026 have claimed that chatbot conversations played a role in teen suicidal behavior. These cases have drawn attention to gaps in chatbot safety systems.
Some companies have responded. Character.ai now bans open-ended chat for users under 18 and requires age verification.
Even so, teens can often find ways around these limits and continue open-ended chats elsewhere.
Doctors suggest parents watch for teens spending more time chatting with AI, pulling back from friends and family, or getting upset when chatbot access is limited. They recommend starting calm, open conversations with teens about how and why they use these tools, rather than banning them outright.
Setting time limits, especially at night, and agreeing on when chatbots should not be used during moments of distress can help. Reviewing how to reach real support, such as a school counselor or the 988 Suicide & Crisis Lifeline, remains part of the guidance doctors are giving families today.
