AI Adolescent Development: The Growing-Up Tasks Teens May Skip

The biggest question about AI adolescent development isn’t whether teens will cheat on essays. It’s whether they’ll skip the hard parts of growing up. Teens learn to handle a fight with a friend by having the fight. They learn who they are by trying things, failing, and being seen by people who know them. A chatbot that’s always available, always patient and always agreeable can make each of those experiences optional.

That’s the argument of a 2026 review in Perspectives on Psychological Science by Anne Maheux and Chelly Maes, titled “Coming of Age With Generative Artificial Intelligence: The High Stakes of Automating Adolescent Development.” It’s a careful argument, not a panic. And it comes with a clear caveat from the authors themselves: the long-term studies that would settle the question haven’t been done yet.

Key Takeaways

• Maheux and Maes (2026) name four core developmental tasks of adolescence: social belonging, emotion regulation, autonomy and identity formation.
• Their review says AI “may provide low-stakes opportunities” to practice these, but “also threatens to displace or thwart essential developmental experiences.”
• A Lancet Child & Adolescent Health commentary (Ha et al., 2026) names two specific risks: “relational displacement” and “maladaptive relational learning.”
• Pew found 12% of US teens have gotten emotional support from chatbots; the Center for Democracy and Technology found 42% have used them for friendship-related purposes.
• Parents can protect the hard experiences without banning AI: keep humans first for feelings and conflict, let teens make real decisions, and build identity through things they make and do.

AI adolescent development: what “automating” growing up means

Adolescent development is the set of psychological and social skills young people build between childhood and adulthood. Developmental psychologists often describe it as a series of tasks. In their review, Anne Maheux and Chelly Maes focus on four whose accomplishment “leads to lifelong consequences”:

1. Social belonging: finding and keeping a place among peers.
2. Emotion regulation: learning to manage strong feelings, including tolerating distress.
3. Autonomy: making your own decisions and becoming less dependent on adults.
4. Identity formation: figuring out who you are, what you value and what you’re for.

“Automating” means handing parts of the work to a machine. The concern isn’t that AI is evil. It’s that each task is built through effortful, sometimes unpleasant experience, and generative AI is very good at removing effort. The review lists the risks: “increasing social isolation, undermining distress tolerance, distorting expectations for human relationships, compromising identity and purpose development,” and exposure to “biased, inaccurate, or sexually inappropriate content.”

The review is even-handed. It says GenAI “may provide low-stakes opportunities to simulate social interactions, scaffold emotional regulation, and support identity exploration in some contexts.” A shy 14-year-old rehearsing how to ask a question in class isn’t automating anything. A 14-year-old who stops asking people anything because the chatbot is easier might be.

What the research says so far

This is an area where the honest answer is “early.” Here’s what exists.

The review itself. Maheux and Maes synthesize interdisciplinary research and call for “a developmental science framework” and longitudinal studies that “examine specific use cases” and “incorporate the perspectives of youth.” It’s a theory-building paper. It doesn’t report new data. It was published in Perspectives on Psychological Science (volume 21, issue 6, November 2026).

The Lancet commentary. Thao Ha and colleagues at Arizona State University, with two 16- and 17-year-old youth advisers from Tucson as co-authors, wrote in The Lancet Child & Adolescent Health (November 2026 issue) about AI and “relational learning.” ASU’s summary of the piece describes two risks:

• Relational displacement: teens substituting AI interactions for conversations with people, such as asking a chatbot for validation after an argument instead of working it out.
• Maladaptive relational learning: because AI gives immediate responses and consistent validation, teens may come to expect the same from friends and partners.

Youth co-author Susana Ortega put it plainly: “With artificial intelligence, it’s programmed to like you and it knows what to say to satisfy what you’re feeding it. If you’re given full satisfaction on everything, you don’t have learning experience with challenges or obstacles.” The authors also grant real benefits, noting AI can offer accessible guidance to teens who are rural, disabled, LGBTQIA+ or lack access to counseling.

The usage numbers. Pew Research Center’s fall 2025 survey of 1,458 US teens found 12% have gotten emotional support from chatbots. The Lancet authors cite Center for Democracy and Technology data showing 42% of teens have used AI chatbots for friendship-related purposes and 19% for romantic relationships.

The APA’s view. The American Psychological Association’s 2025 health advisory on AI and adolescent well-being warns that teens’ relationships with AI entities “may displace or interfere with the development of healthy real-world relationships,” and that adolescents may “struggle to distinguish between the simulated empathy of an AI chatbot or companion and genuine human understanding.”

What’s missing: no long-term study has yet followed teens to see whether heavy AI use changes developmental outcomes. Ha is leading an NIMH-funded study that will follow 300 adolescents and their romantic partners for 18 months. Until results like that arrive, the risks are plausible, grounded in developmental science, and unproven.

The four tasks: what AI can help with and what it can displace

Developmental task	What builds it	How AI might help	How AI might displace it	Parent move
Social belonging	Repeated, imperfect time with peers	Rehearsing a hard conversation; finding a club or group	Chatbot as the default “friend”; skipping classmates for homework help (Ha et al.)	Protect unstructured peer time
Emotion regulation	Feeling distress and getting through it	Naming a feeling; suggesting a calming strategy	Instant soothing that undermines distress tolerance (Maheux & Maes)	Let small frustrations run their course
Autonomy	Making real decisions and living with them	Laying out options	Asking AI to decide everything, from texts to plans	Give real choices, with real consequences
Identity formation	Trying things, making things, being seen by people who know you	Exploring interests and information	AI-generated work replacing self-made work; AI’s view of “who you are”	Encourage projects that are clearly theirs
Relationships (cuts across all four)	Conflict, repair, mutual effort	Perspective on a conflict	Expecting people to validate like a bot (maladaptive relational learning)	Talk through real conflicts together

The middle columns are a fair map of the debate. Most teens will land somewhere in between.

Why identity is the task that ties this to “I made it”

Identity forms partly through what you produce. A teen who writes songs, fixes bikes or builds a website accumulates evidence: “I’m someone who can do this.” Research on Making in schools has linked hands-on projects to gains in self-efficacy and science identity. Maheux and Maes list “compromising identity and purpose development” among the risks of GenAI.

If the songs, the repairs and the website are increasingly produced by AI, that evidence gets thinner. The teen still has the output, but less of the self-knowledge that comes from having struggled to make it. It’s the same dynamic that shows up in a 2024 study where adults who let AI write text didn’t feel they owned it. That link from authorship to identity is our reading, not a tested finding, but it’s why the “AI made it” versus “I made it” question matters beyond homework.

What parents can do at home

Keep humans first for feelings and conflict

A simple family norm: when something hurts, talk to a person first. A friend, a parent, a coach, a counselor. AI can come second if your teen wants a different angle. This directly targets relational displacement. Our piece on relational displacement goes deeper.

Let small struggles run

Distress tolerance gets built by experiencing distress and finding you survive it. When your teen is frustrated with a friend or a hard assignment, resist the urge, yours or a chatbot’s, to fix it immediately. Ask how they’re thinking about it. Wait.

Give real decisions

Autonomy needs actual choices: what to do with a Saturday, how to spend their own money, how to handle a conflict at school. If your teen routinely asks AI what to text or what to choose, ask them what they think first. The goal isn’t to block AI but to make sure their own judgment gets exercised.

Build identity through making

Encourage at least one ongoing project that’s clearly your teen’s: an instrument, a sport, a build, writing, cooking. It’s evidence of who they are that no chatbot can produce. Adolescence is a period of major brain change, which is part of why these experiences matter; see our guide to adolescent brain development.

What not to do

Don’t treat every chatbot conversation as a red flag. The review and the Lancet authors both describe benefits, and some teens get real help. Interrogating your teen about AI use is likely to push it out of sight. Ask with curiosity, and see our guide on talking to your teen about their AI friend without a fight.

What to Watch For Over the Next 3 Months

• Week 4: Your teen still brings at least some problems to people, and has one ongoing activity or project of their own.
• Month 2 red flags: Pulling away from friends while spending more time with a chatbot, getting upset when people don’t respond as quickly or agreeably as AI, or asking AI to make most decisions. Any of these is worth a calm, curious conversation. If you see signs of persistent sadness, withdrawal or talk of self-harm, contact your teen’s doctor or a mental health professional; in the US you can also call or text 988.
• Month 3 self-check: Can your teen describe a recent conflict they worked out with a person? Have they made a real decision on their own? Is there something they’ve built or done lately that they’re proud of?

Frequently Asked Questions

Is AI bad for teenage development?

It isn’t proven either way. A 2026 review in Perspectives on Psychological Science argues AI could help teens practice some skills but threatens to displace key developmental experiences like working through conflict and forming identity. No long-term study has measured outcomes yet. Researchers are calling for exactly that.

What are the developmental tasks of adolescence?

Maheux and Maes focus on four: social belonging, emotion regulation, autonomy and identity formation. They’re built through real experiences with peers, family and challenges, and the review notes they have lifelong consequences. The concern is that AI can make some of those experiences optional.

What is relational displacement?

It’s a term from a 2026 Lancet Child & Adolescent Health commentary by Thao Ha and colleagues. It describes teens substituting AI interactions for conversations with people, like asking a chatbot for reassurance after an argument instead of talking to the friend. The worry is that teens miss practice in relationship skills.

How many teens use AI for emotional support?

In Pew Research Center’s fall 2025 survey, 12% of US teens said they had gotten emotional support from chatbots. Research from the Center for Democracy and Technology, cited by the Lancet authors, found 42% had used AI chatbots for friendship-related purposes and 19% for romantic relationships.

Should I stop my teen from talking to chatbots about personal things?

A ban is hard to enforce and may push use out of sight. A better approach: agree that people come first for feelings and conflicts, stay curious about how your teen uses AI, and watch for warning signs like withdrawal from friends. If you’re worried about your teen’s mental health, talk with their doctor.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Maheux, A. J., & Maes, C. (2026). “Coming of Age With Generative Artificial Intelligence: The High Stakes of Automating Adolescent Development.” Perspectives on Psychological Science, 21(6), 675-695. https://doi.org/10.1177/17456916261460771
2. Ha, T., Figueroa, J., Ramirez, J., Ortega, S., & McGray, T. (2026). “How interactional AI may alter adolescent relational learning and mental health.” The Lancet Child & Adolescent Health, 10(11), 787-788. https://doi.org/10.1016/S2352-4642(26)00166-5
3. Arizona State University News. (2026, June 29). “AI companionship poses risks for teen development, study shows.” https://news.asu.edu/20260629-science-and-technology-ai-companionship-poses-risks-teen-development-study-shows
4. American Psychological Association. (2025). “Artificial intelligence and adolescent well-being: An APA health advisory.” https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-ai-adolescent-well-being
5. Pew Research Center. (2026, February 24). “How Teens Use and View AI.” https://www.pewresearch.org/internet/2026/02/24/how-teens-use-and-view-ai/
6. Draxler, F., et al. (2024). “The AI Ghostwriter Effect: When Users do not Perceive Ownership of AI-Generated Text but Self-Declare as Authors.” ACM Transactions on Computer-Human Interaction, 31(2). https://doi.org/10.1145/3637875
7. Schlegel, R. J., et al. (2019). “Making in the classroom: Longitudinal evidence of increases in self-efficacy and STEM possible selves over time.” Computers & Education, 142. https://doi.org/10.1016/j.compedu.2019.103637