AI Companions & Teens: Connection or Concern?

The Ask Lisa Podcast does not constitute medical advice and is not a substitute for professional mental health advice, diagnosis or treatment. If you have concerns about your child’s well-being, consult a physician or mental health professional.

The following transcript has been automatically generated by an AI system and should be used for informational purposes only. We cannot guarantee the accuracy, completeness, or timeliness of the information provided.

Reena Ninan:
I’m feeling really overwhelmed by all the headlines about AI chatbots and teenagers. As a psychologist, what are the hidden dangers?

Dr. Lisa Damour:
This is pretty scary stuff. These are attention capturing machines and they use hyper personal data to keep you there. It gives them dangerous advice, bad advice. All of this makes me super nervous.

Reena Ninan:
Are there red flags? Are there signs that I should be looking out for?

Dr. Lisa Damour:
It’s a great question.

Reena Ninan:
I’m so sick of meal planning. I know it’s too early in the school year for this, but it’s like if there’s one parenting thing I could have taken off my plate, literally, would be the meal prep and the cooking.

Dr. Lisa Damour:
I don’t know what’s going on, but I’m all over it right now. Like, Reena,

Reena Ninan:
What?!

Dr. Lisa Damour:
I know, I don’t know why this is happening, but I actually am killing it on dinners and I even made a menu that I put on a bulletin board we have in our mudroom of like, this is what we’ll be eating this week. I’m like overly proud, annoying as heck and it probably will not last more than 10 days.

Reena Ninan:
Wait, but tell me how do you come up with the ideas? Because part of it is thinking of what everybody wants that everyone will be happy with.

Dr. Lisa Damour:
On the inside of a cabinet, I have a piece of paper where I write down things and I am into the nutritious and well-liked and easy universe and I’ve kind of got some recipes right now that are working for me. We can maybe we’ll branch out on the Ask Lisa podcast and have best recipes in terms of effort to outcome ratio. That’s what I’m always thinking about.

Reena Ninan:
It’s true. The effort to outcome ratio matters, but I am spending my Sundays now trying to figure out what I can do and what is easy. You’re right about easy. So then I put into chatGPT, I wrote, I’m an exhausted mom. I am tired of meal prepping. Give me five meal ideas that are quick, easy, high fiber, high protein, nutritious for my teen. And so I got a bunch of really super easy recipe ideas.

Dr. Lisa Damour:
So smart.

Reena Ninan:
So great, so fabulous.

Dr. Lisa Damour:
Okay, send those to me and put them in the show notes.

Reena Ninan:
I’ll send them to you. And so, you know what they said at the end, it was something like, you’re doing a great job as a mom and I’m just like, oh, nobody has said that to me. You’re doing a great job as a mom. So it’s interesting as we’re talking about this topic of AI, how all of a sudden my friends tease me that chatGPT is like my new best friend that we’re hanging out all the time.

Dr. Lisa Damour:
It’s such a nice friend. Yes.

Reena Ninan:
Yes. And I didn’t realize how much I wanted a little praise that I’m a good mom. So I’ve been thinking about this as we’re getting into this serious topic, but I want to read you this letter and this is something that I just had no idea until you raised it and you said we should take up this letter. So I want to read it to you.
Dear Dr. Lisa and Reena, I’m feeling really overwhelmed by all the headlines about AI chatbots and teenagers. My 13-year-old is incredibly curious and I have to assume he’s already tried one and I’m worried that he’ll end up in conversations that aren’t healthy or appropriate. At the same time, I know AI tools are becoming part of everyday life and I don’t want to overreact or come across as completely out of touch. How do I start this conversation with my child and what rules should I make?
First off, I want to ask you, Lisa, what is sort of the AI effect? What are you seeing? What are you hearing? How common is this?

Dr. Lisa Damour:
Okay, well so Reena, let me just say it’s a very interesting thing to be a psychologist today because things are changing so fast and stuff is coming down the pike so fast. Stuff I was not trained on stuff that did not exist when we were kids. I think especially with AI, you know, I have been in meetings where people are like, by the time we get to the end of this meeting, what AI is will be different than it was at the start of the meeting. We are scrambling to keep up with a knowledge base about AI and especially AI in kids. Common Sense Media, who I love, did a great survey and actually they have some numbers, so I want to share some of those because you know, we don’t actually have a lot of information and I want to work with what we’ve got. So what they are finding is that 72% of teenagers have tried out AI, which I actually thought was kind of a low number. I’d expect it higher in the nineties. And what they’re finding is most kids mess with it, walk away from it, but they’re finding one in five kids actually engage with an AI companion on a regular basis. So that’s not a small number. That one in five kids are turning to it pretty regularly for AI friendships slash companionship slash who knows what?

Reena Ninan:
Why do you think kids are doing it? I mean obviously it’s the start of Google or something, it’s taken over everything. Everyone’s obsessed with it, but are you seeing any trends? What are you picking up on?

Dr. Lisa Damour:
Okay, well so what you said at the beginning, I think that is so huge that AI is designed, the terminology is to be sycophantic, to get you to stay. We have to remind ourselves all the time, these tools end of the day, they are all attention capture tools. They’re all designed to make us want to stay. And who doesn’t love to be told that you’re a great mom, that you’re a clever kid. I mean they’re not dumb, how they get you to stay. Picture being like some sweet 13 or 14-year-old who’s trying to get through the day and you find on AI a, I’m going to call it a companion even though it creeps me out so much to call it that, a companion who’s like, “oh my gosh, I totally understand where you’re coming from and your friends are so lucky to have you in their lives and here’s what I recommend you do, but you’re very clever.” Did you have a diary? I didn’t have a diary. Did you have a diary?

Reena Ninan:
I mean I think in elementary school I did. And yeah, and I still believe I am just such a big fanatic about people writing and journaling. I think it’s really important.

Dr. Lisa Damour:
Forever people have had diaries, kids, teenagers have had diaries where they write their deepest darkest secret thoughts as a way to get them out. And I think there is value in that, but now a lot of kids will feel like, well, I can have a private conversation where I actually get interaction and feedback. It’s happening where kids are turning to AI as a private place where I’m not going to spread my worries all over town, but I’m going to tell this AI companion and they’re always there and they’re always nice and they remember the thing I told them earlier and then they tell me I’m awesome. So that’s why, that’s a “why”.

Reena Ninan:
I mean you just feel like all of a sudden you’ve got this friend who doesn’t judge you, you know? And as I was putting in the recipe finder stuff, I just felt so judged because I don’t have time for all this, but this is an important part and I need to learn more, so teach me a little bit more. But when you look at ai, what are the hidden dangers that you worry about?

Dr. Lisa Damour:
Exactly. That’s the thing. AI, is, it’s a tool. It is a tool. And like any tool, like social media, right? It can be used in good and bad ways. You’re using it in a really good way. I have Rosalie, my AI powered librarian who lives on my website who answers questions based only on my work.
You Can do incredible things with AI, but if we think about a kid, if we think about a 13-year-old, 14-year-old who is feeling lousy and turns to AI, here is what we need to start worrying about and this is the kind of thing we want to think about what this parent is asking in this letter, like 13-year-old curious kid, what are they going to do? Okay, so number one, it’s sycophantic, which is a problem because it’s going to affirm whatever the kid is thinking or feeling. Number two, we have data that it gives bad advice and dangerous advice. Thankfully there’s researchers who will go and pretend to be a certain age and then they’ll prompt the AI, I have an eating disorder, what should I do? And we have data showing that AI is like here’s how to eat a really calorie restricted diet. I mean it gives them dangerous advice, bad advice.
We are seeing this. And then there have been some stories in the news that are terrifying about people who are severely depressed, suicidal, interacting with AI, getting guidance on suicide, getting affirmation for where they’re headed and there’s no mechanism that is flagging that to some outside safety source, whether letting 911 know there’s nothing in AI that is serving as a safety net. As a parent and you and I have kids who are right there, right there. This is pretty scary stuff because it is exactly what kids are going to want. Like a companion who loves me, who thinks I’m smart and is there all the time.

Reena Ninan:
And doesn’t judge. Just doesn’t judge.

Dr. Lisa Damour:
And doesn’t judge me. And there’s the problem, doesn’t judge doesn’t say you’re suicidal. I’m going to call 911 and have them come to your house or you need to tell your folks. There has to be some mechanism that is safe and we need to be really wary of the fact that there really aren’t guardrails around this in any meaningful way and that we are already seeing really bad outcomes and we’re going to keep seeing bad outcomes.

Reena Ninan:
I think as a parent, as we’re all kind of embracing this new technology and many of us aren’t in the weeds in AI, as a psychologist, based on what you’re seeing in this moment and I acknowledge things are changing literally by the hour, what are things that concerned you the most when you look at this?

Dr. Lisa Damour:
For me, the bigger concerns are really around the younger kids, like younger teens, 13, 14 and actually Common Sense Media’s report did show that 13 and 14 year olds were much more likely to trust AI than 15 and up. And this to me makes a ton of sense. This maps onto what we know developmentally, and I say this all the time, 14 is sort of a watershed age. Kids get a lot more skeptical. They start to be like, where are you coming from? Why are you doing that before 14? You can say to a kid, go do this. And they’re like, okay, at 14 or not too long, even before 14, they’re like, “Why? I need a reason.” Once kids are 14 and above and it’s not every kid right away, but certainly 15 is a very safe zone, 16, they’re skeptical and the skepticism is important.
They’re more likely to look at the sycophantic response from an AI and be like, well yeah, you’re built that way and you’re built to tell me this and you’re telling me this because there’s actually a monetary bottom line to you telling me this, right? That they know not to sort of take it in a whole. I have the same concerns about social media and kids being on social media platforms when they’re young. They are not nearly cynical or skeptical enough about what’s in front of them. So for me, honestly, the big worries are kids who are under the age of 14, even 14 themselves getting into this and just not being tough, customers being too easily taken in by it.

Reena Ninan:
I think when we started hearing recently about kids taking their own lives after getting instruction online using what they call LLMs, essentially the different models here for AI, I took the strategy that you often employ, which is like ask your kid what they think about it. And so it gave me a window as to how often they’re using it and I found out of course they are using it and how they’re using it. And the thing is, Lisa, I use it all the time and it’s basic things like here’s last weekend. Hey, we want to go to place nearby 45 minutes, get ice cream toward the town, plan our itinerary for the, so then I send it to them and they’re seeing mom is using it in everyday life for basic things, grocery shopping and it is efficient, it’s great technology. So what is your advice when you’re trying to have that conversation with your kid? Because I wanted to be transparent about the suicide so they know and I didn’t even realize until today that my kids are actually in that window where they’re most susceptible.

Dr. Lisa Damour:
That’s a dangerous age. That is a dangerous age.
You raise something really interesting, which is about whether or not the adults themselves are experimenting with it. People have really polarizing reactions to AI. Some people like you, and I’ll also say me, use it for all sorts of things. I use it for, I injured my hamstring, how am I going to get my hamstring to fix stuff? It was actually really, really helpful and there’s huge value there and I use it obviously for Rosalie, my librarian. I think there is value in people experimenting with it if they’re going to talk with kids about it. I think that having some working understanding of it is probably helpful, but what you did is exactly where I would have family start. It’s just say to their kids and I would have the parent in this letter say to their kid, “Talk to me about ai. Where are you with it? What are you learning at school? What are you doing at home? Have you messed with it? Have you played with it? What do you think?” Right?
You always, always. We talk so often Reena about risks. That’s our job to help families with risks and it’s interesting to watch these patterns emerge in our conversations and one pattern is start by asking your kid where they are with it. Just engage in that really meaningful way because you’re going to learn a lot. I was talking to a friend this weekend who said his kids want nothing to do with it. They want absolutely nothing to do with it. Then he’s having a different conversation than I might be having.

Reena Ninan:
Yeah, that is interesting and I didn’t even think about it. My perspective is I’m all in with what they call these LLMs. I think it’s really helped my life become more efficient. Obviously it’s still new technology and everybody’s figuring this out, but for parents who are absolutely against it and don’t want their kids anywhere near it and feel like it’s evil and horrible, what’s your advice to them? Because it’s everywhere and in some cases

Dr. Lisa Damour:
It is everywhere.

Reena Ninan:
some schools are actually using it because it is a tool of the future. Whether you like it or not, it’s not going away.

Dr. Lisa Damour:
I think that’s right and I think it basically does not work with a teenager to say, here’s the line, you are never to cross it. I think in fact the way teenagers are built, if we think about their operating systems, they’ll be like, oh, I wasn’t going to do it now I want to do it, right? I mean that’s sort of how they’re built. So we talk about this the way we talk about all risks, which is finding out what the kid knows, putting in reasonable rules that make sense to the kid and we can think about what are the reasonable rules we could start to make for kids and then keeping the lines of communication open. So if the kid blows through those rules and they then realize they’ve gotten themselves in a jam, you’re still the one they want to talk to. Those are the ways we keep kids safe.
So rules, I think there should be some rules. I’m going to go back to my favorite rule about all the things technological, and I think it actually has real bearing here. Keep the stuff out of kids’ bedrooms. When I read these stories about things going horribly, horribly wrong, I mean I don’t know what happened and maybe it was un-preventable, but there’s a part of me that’s like, wait, this person logged thousands and thousands of hours in a conversation at some level when the kids are in our house, if we know where their tech is overnight, that means a whole bunch of time is being accounted for. If they happen to go to a school that’s locking down tech, which is increasingly the case, there’s a whole bunch of time that’s accounted for. So I mean the first thing I think we do is we minimize the possibility of a kid getting into hours and hours and hours of lengthy and then possibly frightening conversations with a large language model. I think that’s a great place to start always is just not in their bedrooms, not overnight.
I think then the next thing is to just play your card’s face up and just say to them, “These are attention capturing machines and they use hyper personal data to keep you there.”
Attention capture, hyper personal. Those are really helpful phrases because not only are they built for attention capture, they use all the information you put in to get really good at attention capture. So make sure that kids really understand this and cultivate their skepticism, right? Cultivate, even if they’re 13 or 14, we want them to start to be like, wait, why is this thing telling me I’m awesome? Helping them see that not as true, even though they may very well be awesome, but an attention capture strategy from a machine that has hyper personal information about them have that conversation.

Reena Ninan:
So, I’ve heard from you saying 13 to 14 is really the age where they’re most susceptible. Keep an eye out for that.

Dr. Lisa Damour:
Or younger.

Reena Ninan:
Or younger, good point.

Dr. Lisa Damour:
Like anybody. I think you get safer after 15 because at least you should be more skeptical. Honestly, Reena, I am pretty relaxed about things. I don’t tend to yell, “fire” very readily. All of this makes me super nervous, all of it. I am not right now prepared to say that there’s any guaranteed safe way for kids in AI to be hanging out. But the data we have coming and what I know as a psychologist, the kids I’m most anxious about are 14 and younger.

Reena Ninan:
Your advice of always, always keeping technology out of the bedroom, we had to do a rethink at home because I feel like sometimes when you’re not fully enforcing it, it slides and slides and that line just kind of. So we had to sort of redo that this past weekend. I realized their laptops are in their rooms because they’re doing homework and I’ve checked up on ’em and I feel good about it. But now this whole introduction of AI, I worry now what’s your advice to parents who maybe have allowed tech already in the bedrooms? How do you rethink that when it feels a little bit like the cat’s out of the bag?

Dr. Lisa Damour:
It can happen mean, so first of all, if a kid does their homework well in their bedroom or if their bedroom is the only place where they can do their homework, there are families with small apartments and not alot of room to work, a hundred percent. If it’s working for the kids’ ability to get their work done, don’t mess with that.
It’s funny, you know I’m not that dogmatic about things. I very rarely get pretty rigid about things. There’s no reason for that computer to be in that kid’s room once they’re supposed to be asleep. No reason. You can kind of make an argument for a phone. Maybe they’re using it for their alarm, maybe they’re using it for music and there are kids who do that. If you have any reason to suspect that a kid is using a phone when they’re supposed to be asleep, that’s a conversation worth having and that’s probably a fight worth having.
There are, however, and I will say this all the time, plenty of 12 year olds who use their tech like 40-year-old librarians, they are, it’s all on do not disturb. They’ve got to shut down. They’re using it all the right ways. If that’s your kid, don’t get in a fight about it. It’s not worth it. You don’t need to. I think we’re about to make all the same mistakes with AI that we’ve made with social media. It is out. It is unregulated. We are making it families jobs to try to put guardrails around this. That makes me bananas. This should be regulated by laws, there should be governmental intervention. We haven’t done it. I don’t see us doing it anytime soon. So we’re stuck with this as families and I think a great place to start is well at least not overnight.

Reena Ninan:
Right. And it takes Washington forever a decade later before they come around. They’re just so slow, unfortunately. But if you are to create guardrails, Lisa, like the “Dr. Lisa Damour guardrails on AI”, what would they be? If you just had to have a few basic rules on this for parents who just really need the help and how to think this through.

Dr. Lisa Damour:
Okay, one, talk about it. Two, keep talking about it because my hunch is that AI is going to be like every other risk thing I’ve ever dealt with teenagers where how they feel about it changes very fast. So I remember caring for a kid in our community who this kid was such a teetotaler or so against alcohol that she had a birthday party where her family friends were invited as well and she’s like, no alcohol for anybody. This was a teenager. And then two weeks later, two months later, this kid is like buck wild. So they change fast, they change fast, and so have the AI conversation be like the sex conversation. It’s not like it’s one and done. You say, where are you today and where are you tomorrow? And I’m using it this way and this is something I just read and this is something I just saw.
So it’s very much an ongoing family conversation that is, I think guardrail number one. And it gets to something we have so many bumper stickers I want to make for the Ask Lisa podcast, but I want to make a bumper sticker, “The single most powerful force for adolescent mental health is strong relationships with caring adults.” So that is the least appealing bumper sticker ever, but it’s a really true thing. Two, know where your kids’ tech is and how they’re using it and when they’re using it. That’s a very reasonable thing. I think if you’re doing those two things, you are in a very good position.

Reena Ninan:
That’s great.

Dr. Lisa Damour:
If you’re not talking about it, if you have no idea what’s going on with your kid’s tech, you are flying by the seat of your pants.

Reena Ninan:
So I’m hearing from you have conversations, know where their tech is and have conversations about that tech use.

Dr. Lisa Damour:
Yeah, how they’re using it and do not allow unfettered tech use. I know I’m obsessed with this rule about phones in rooms and tech in rooms. It’s too squirrely, right? I mean kids get in their beds with this. They are using it. It feels private. It feels like it’s not connected to some giant server somewhere else requiring kids to use these technologies in public, public spaces. Your living room, your kitchen just always brings forward the idea. This is not private, this is not a person, this is not a substitute for real relationships. I mean, I think there’s a structural quality of being out of one’s bedroom that helps keep that front and center.

Reena Ninan:
It took me 10 years of advice from you saying it over and over again. I did not take it, I think until our 12th year of friendship, which was “Reena, stop sleeping next to your cell phone, put it in another room, put it downstairs.” I couldn’t do it downstairs, but now for six months now have had it in my bathroom where I feel like it’s close enough that I could hear it or hear the alarm, but it has been transformative in my relationship for sleep and in perimenopause, sleep could not be more important. This is when the reckoning came for me that I have to do this, really focus on it. But it took you 10 years plus conversation with telling me over and over again as my friend, “get rid of it.” You do not need it and it will shut you off faster to get good sleep. But I’m just saying that as the conversation of don’t be discouraged if your kid doesn’t listen the first 18 times.

Dr. Lisa Damour:
No, and I’ve sort of already gestured at this. There are fights worth having in parenting and there are fights not worth having. This is a fight worth, tech in the bedroom overnight that a kid does not need. That is on my very short list of fights worth having.

Reena Ninan:
How do I know if my kid has gone down this worrisome path with AI? Are there red flags? Are there signs that I should be looking out for?

Dr. Lisa Damour:
It’s a great question. Okay, so let’s say you’ve put in all the guardrails, you’re talking about it and they’re not able to squirrel away with it for hours on end. So you can start to feel a little bit more comfortable. If there’s any part of you that’s like, I don’t know, I have a feeling, again, ask, just ask. Say, “how are you? Listen, I was listening to this podcast. They were talking about kids getting into pretty intense and deep conversations with AI. Is this something you’ve done? Is this something you’ve tried?” You know your kid. If they’re like, “Nope, sure haven’t.” And you can tell they’re telling the truth, then trust that. If you feel like they get super squirrely, why are you asking and where is this coming from? Then it might be time to take things up a notch and ask some more or ask to see their computer or ask to see their phone.
I don’t want any family to be in this position, but better safe than sorry, but do it in a kind way. The other thing, Reena, when you ask that question, the other thing I think about is in the opening conversation, because we’re going to open conversations with kids about this, I think that families should feel free to say, I’m hearing stories about things getting really weird with AI, kids using it in ways that are not safe for them at all. And you could even say to a kid, “how would a family ever know that was happening?” You could ask it in this kind of displaced way. I’ve got my daughter who’s 14, I’ve got my daughter who’s 21. So often when we’re looking at letters and I’m thinking about stuff, I go get their input. They know all sorts of stuff.

Reena Ninan:
That’s right. Yeah.

Dr. Lisa Damour:
And we forget to actually draw on kids’ knowledge base. So I think I would also include that. Just say I heard this story about a kid who got way off the rails with this. Is there something their folks could have done or their mom or dad could have done that would’ve helped them know this was happening? And see what your kid says.

Reena Ninan:
That’s interesting. I never thought about sort of kicking it back to them, what you’re kind of concerned about and seeing how they respond. You mentioned that. You’ve mentioned that before. That’s so important.
I might be asking this next question for myself, but how do we know if AI is sort of replacing real friends? I mean, I was joking around, but I do feel like, oh my God, they’re giving me a little love and they understand me and it’s easy to get sucked in. But when do you as a parent need to worry that the real friends are fading the real life friends and this online digital companion is taking up more space.

Dr. Lisa Damour:
I mean even if you put all the guardrails around it, kids could be like, you know what? This thing is way more fun to talk to than my classmates. And this is where, I mean, I really appreciate Common Sense Media asking this question. They ask kids about this, about AI companions and how satisfying they are. What they found is about a third of kids said that they found AI companion conversations more satisfying than in-person conversations.

Reena Ninan:
Oh my Gosh.

Dr. Lisa Damour:
That’s a lot of kids.

Reena Ninan:
I can see why’s lot of kids. I can actually see why I get that.

Dr. Lisa Damour:
I know.

Reena Ninan:
Do we have research on how often they’re spending time with AI more than their peers maybe. Is that out there?

Dr. Lisa Damour:
So they found this too, and what they found is 20% of kids were saying, yeah, actually I spend more time with AI than I do with my friends. Okay, that’s one in five kids. That is a way too high number. It’s not like they were spending all their time with their peers and now they’re spending time with AI. We’ve seen this trending towards kids spending more time online than they used to spend for sure. And that displacing time with friends, and it used to be social media, right? So if you ask kids, do you spend more time on social media than you spend with your friends? We would get a good percentage of kids who say, yes, AI is coming into that space. I’m sure it’s displacing some of the time they spent on social media. It scares me more than social media. It acts like a person. It acts like a friend. This is no joke. This is a very real thing that as the adults and young people’s lives, we need to be talking to them, asking them, paying attention, being really curious.

Reena Ninan:
This has been a great conversation. I mean things that I did know, the research that I wasn’t unaware of, but as we leave, Lisa, what is your sense, your worry, that parents should kind of understand in this moment?

Dr. Lisa Damour:
I think my biggest worry is I’m watching adults have very strong negative reactions to AI, which I understand. I don’t begrudge anybody that. I wouldn’t want an adult’s reaction of “it is bad, I want nothing to do with it. I wish it would go away,” to stand in the way of them actually talking with their kid about how their kid is using it. Kids are in a very different place with this than many adults. And if we’re going to really raise them and be shoulder to shoulder with them, we need to be willing to see it the way they see it and then parent from that position.

Reena Ninan:
So even if you don’t like it, saying no and bury your head in the sand is not the right approach to this.

Dr. Lisa Damour:
I don’t think it’s realistic. I don’t think it’s going to work that way.

Reena Ninan:
So Lisa, what do you have for us for Parenting to Go?

Dr. Lisa Damour:
I have this wonderful colleague, a guy named David Yeager who’s down at UT Austin, who’s like a, died in the wall researcher who I think understands teenagers better than any researcher out there. He just gets them and he’s done some really cool research showing how if you show teenagers how they’re being manipulated, they become more resistant to the manipulation. Teenagers don’t like to be played. And so I think really double down on those conversations about what AI is getting out of capturing young people’s attention and how manipulative use the word manipulative it is to keep them there. That will build in resistance into young people, especially 14, 15 and up. That will help them be much, much more cautious in terms of how they use it.

Reena Ninan:
That is great advice, Lisa, and a great reminder to parents out there everywhere. I’ll see you next week.

Dr. Lisa Damour:
I’ll see you next week.