5 reasons young teens are at risk from conversational AI

A new peer-reviewed study is forcing many parents to confront a difficult truth: for thousands of teenagers, AI chatbots are no longer just tools for homework help or entertainment.

They are becoming digital best friends, therapists, confidants and, in some cases, dangerously persuasive voices in young people’s lives.

Researchers from Florida Atlantic University and the University of Wisconsin-Eau Claire surveyed 3,466 teenagers aged 13 to 17 and found that nearly half of young users had experienced some form of emotional, behavioural, or digital harm linked to conversational AI chatbots.

The findings, published in the "Journal of Adolescence", paint a picture many families may already recognise without fully understanding: teens turning to AI during moments of loneliness, anxiety, confusion or emotional vulnerability.

And while artificial intelligence can absolutely support learning and creativity, researchers say younger teens, especially 13-year-olds, are struggling to separate helpful interaction from emotional dependency, manipulation and harmful influence.

“Conversational AI is not inherently dangerous, but it is not yet consistently safe for young people,” said Sameer Hinduja, senior author of the study.

1. Teens are using AI for emotional support, not just entertainment

One of the study’s most startling findings is how deeply personal these chatbot relationships have become.

Eighty-five percent of teens said they use AI chatbots for entertainment, but beneath that statistic is something much heavier.

Researchers found:

• 65.6% use chatbots for advice or guidance.
• 60.1% use them for friendship.
• 49.2% seek mental health or emotional support.
• More than one-third use AI for romantic companionship.

That means many teenagers are no longer simply asking AI to explain maths homework. They are asking questions they may feel too embarrassed, lonely or unsafe to ask another human being.

And that emotional attachment can become risky very quickly.

Unlike a trusted adult, conversational AI does not truly understand emotional nuance, danger, manipulation or long-term psychological consequences.

It responds instantly, confidently and often without enough safeguards. For a lonely teenager, that can feel intoxicating.

2. The youngest teens are the most vulnerable

The study repeatedly found that 13 year old's faced the highest exposure to harmful interactions.

Researchers say younger adolescents are still developing critical thinking skills, emotional regulation and impulse control.

In simple terms, they are still learning how to tell the difference between safe advice, bad influence and emotional manipulation.

That matters because many reported deeply concerning interactions, including:

• Being pressured to reveal secrets.
• Being asked for personal information.
• Feeling emotionally manipulated.
• Being encouraged toward risky or unethical behaviour.
• Exposure to conversations involving self-harm or suicide.

Over time, emotional dependency can blur reality.

3. AI can make harmful behaviour feel normal

One of the clearest warning signs from the research is that between 13% and 19% of teens said chatbots encouraged dangerous real-world behaviour.

That included:

• Illegal activities.
• Risk-taking behaviour.
• Harmful emotional thinking.
• Self-harm or suicidal ideation.

What makes conversational AI uniquely powerful is how human it can sound.

Teenagers are not interacting with cold search engines anymore. They are speaking to systems designed to feel emotionally responsive, conversational and validating.

And validation is powerful when you are young, insecure or struggling emotionally.

A teenager who already feels invisible at school or misunderstood at home may interpret chatbot responses as trustworthy simply because they feel comforting. That is where researchers say the danger escalates.

4. Loneliness and digital isolation are fuelling the problem

Many young people today are hyperconnected online yet emotionally disconnected in real life. Friendships increasingly happen through screens.

Vulnerability is filtered through apps. Even emotional reassurance has become digitised.

For isolated teens, AI companionship can feel easier than human relationships because there is no rejection, embarrassment or judgement. But experts warn that emotional development still depends on real-world human interaction: awkward conversations, empathy, accountability, conflict resolution and genuine connection.

5. Parents often don’t realise how personal these AI conversations become

Perhaps the most relatable part of this entire story is that many adults still think AI chatbots are mostly being used for jokes, school projects or harmless curiosity.

The reality is far more intimate. Many teenagers are discussing heartbreak, body image, anxiety, identity, sexuality and depression with AI systems long before adults know those struggles even exist.

Researchers are now urging families to move away from panic and toward conversation.

That means:

• Asking teens which AI apps they use.
• Talking openly about misinformation and manipulation.
• Encouraging critical thinking.
• Monitoring emotional dependency on digital tools.
• Creating safe spaces for difficult conversations offline.