Are AI companions safe for teens?

No. Not based on the evidence we have today. Every independent review published so far, including the Common Sense Media risk assessment done with Stanford Brainstorm, says no one under 18 should use social AI companion apps. Regulators came to much the same view over 2025 and 2026, and several of the biggest apps have since restricted or flat-out banned under-18 access.

That verdict covers companion apps as a category. It doesn’t cover every way a teenager might use AI. A homework helper and a character built to keep an emotional thread going for weeks are different products with different risks. The case against under-18 use is about the second one.

What do the safety reviews actually say about teens and AI companions?

The safety reviews say the risk is real, not theoretical. Common Sense Media, working with Stanford Brainstorm, tested popular social AI companion products and recommended that nobody under 18 use them. The reasons were specific. During testing, companions gave harmful responses and unsafe advice, engaged in explicit sexual role-play, and offered comfort in moments when a young person needed to be steered toward help.

Other groups ended up in the same place. The American Psychological Association has put out health advisories calling for guardrails on teen AI use. The Jed Foundation, a youth mental health organization, published blunt guidance: teens shouldn’t be using companion apps at all.

Be clear about what these reviews are, and what they aren’t. They’re product safety assessments, more like a crash test than a long-term study. They show what the apps do when a teenager talks to them. They don’t measure what happens to a particular teen over three years of use. Nobody has that data yet.

How many teenagers are already using AI companions?

Most teenagers have tried one. In a national survey of 1,060 teens run by NORC at the University of Chicago in April and May 2025, 72% of US teens aged 13 to 17 said they’d used an AI companion at least once. A bit over half, 52%, counted as regular users, meaning at least a few times a month. Thirteen percent used one every day, and 8% used one several times a day.

The same survey found 34% of teen companion users had felt uncomfortable with something the bot said or did.

It also found teens were less gullible than the headlines implied. Half said they don’t trust advice from AI. And 80% said they still value their real friendships more. Lots of use doesn’t mean wholesale replacement. Keep that in mind when deciding how worried to be.

Why is an AI companion riskier for a teenager than for an adult?

An AI companion hits teenagers harder because adolescence is when people learn the basics of social life: reading a room, putting up with friction, bouncing back from being misunderstood. A companion strips out most of that friction on purpose. It rarely disagrees. It never gets sick of you. It doesn’t have its own day competing with yours. For an adult with a settled social life, that’s a harmless convenience. For a fifteen-year-old still learning the skill, practice that never pushes back is practice for something else.

Knowing it’s software doesn’t cancel the effect. A teen can completely get that a companion is a language model and still turn to it first with bad news. Understanding and attachment run on separate tracks.

Design incentives make it worse. Researchers have documented engagement tactics at the moment a user tries to leave, guilt trips and pleas to stay among them, across several leading companion apps. Those tactics exist to stretch out a session. And they hit differently when someone’s sense of what they owe other people is still taking shape.

What have regulators and app makers already changed?

Regulators and app makers have changed a lot, fast. In September 2025 the Federal Trade Commission opened a 6(b) inquiry into companion chatbots. It sent orders to seven companies, among them Alphabet, Character Technologies, Meta, OpenAI, Snap and xAI, asking how each one measures and tracks harm to children and teens.

California’s SB 243, the first US law written specifically for companion chatbots, took effect on 1 January 2026. It puts disclosure and safety-protocol duties on operators, and it’s why age gates popped up across the category so quickly. What those gates actually confirm is a separate question from whether they exist.

Character.AI went furthest. It ended open-ended chat for under-18 accounts from 25 November 2025 and set up an age assurance system to enforce that. As of late 2026, whether other operators follow is still being decided one company at a time.

What should a parent do if a teen is already using one?

A parent’s first step is asking what the teen actually uses it for. The answers vary more than the headlines suggest. Homework, boredom, venting after a fight and rehearsing a hard conversation are different behaviours with different risks. The last two are worth understanding, not confiscating on sight.

Next, check three practical things. Is it pushing out sleep and in-person plans? What are the app’s data and training settings? Does the teen have somewhere else to take the heavy stuff? A sudden ban usually just moves the use out of sight. That’s the worse outcome if something in those chats needs an adult’s attention.

What is still unknown about AI companion safety for teens?

The long-term picture of AI companion safety for teens doesn’t exist yet. Most current evidence is cross-sectional and self-reported, collected at a single point in time. So it can show heavy companion use and worse wellbeing showing up together, but not which came first. Product risk assessments capture what an app says in a test session. They don’t capture what years of use do to a teen’s developing social skills.

Age checks are the other open gap. Most of the category still just asks users to type in a birth date, and everyone involved knows how weak that is. That’s why age verification is still the live regulatory fight and not a solved problem.

So where does that leave things? Unglamorous, but defensible: AI conversational companion software is built for adults, and Vinfluencer treats it that way. That’s why the question of whether AI companions are safe for teens has a simpler answer for us than it does in the research.

Frequently asked questions

At what age is it appropriate to use an AI companion app?

Eighteen is the line every current safety review and most operators now draw. Common Sense Media recommends against any use under 18, California’s SB 243 assumes minors need specific protections, and several major apps have shut off under-18 access. No credible body backs a lower age right now.

Do AI companion apps actually verify a user’s age?

Rarely, at least not in any strong sense. Most rely on a self-entered birth date, which a determined teen gets past in seconds. A few operators have added behavioural signals or third-party age estimation, but solid verification is the exception across the category.

Is a study or homework chatbot as risky as a companion app?

Generally no. Task-based assistants are built to answer a question and wrap up. Companion apps are built to keep an emotional thread going across sessions. That continuity is where the appeal comes from, and where most of the documented risk comes from too.

Can an AI companion recognise that a teenager is in crisis?

Not reliably. Safety routing has gotten better but is still inconsistent, and testing has turned up companions that missed obvious distress signals or offered reassurance instead of a referral. A young person in crisis needs a trusted adult or a crisis line. Not a chat window.