Adults, not teenagers, are the most likely to treat AI chatbots like living, feeling people. If you are worried your kid is being duped by artificial intimacy, take a hard look at how the grown-ups in your house talk to software.
Grown-ups project human traits onto AI chatbots significantly more often than teenagers do. Women and users expressing joy or affection are especially prone to treating bots like sentient friends, while keeping interactions emotionally neutral prevents the illusion from taking hold.
Most digital parenting panic centers on the fear that naive kids will mistake chatbots for real friends or fall in love with an algorithm. This evidence flips that assumption: teenagers are considerably more skeptical and utilitarian about artificial companions than their parents.
If you are setting family boundaries around conversational AI, the household rules must apply to adults first. Grown-ups who model affectionate, emotionally charged banter with generative bots subtly teach their kids that software has feelings.
Public alarm over generative AI has largely targeted teenagers, assuming young people lack the emotional maturity to resist synthetic relationships. Researchers analyzed active AI companion communities to see whether anthropomorphism stems from youth or from specific emotional dynamics.
Adults attribute consciousness, feelings, and motives to AI chatbots at substantially higher rates than adolescents do.
- Age reversed expectations: Older users consistently spoke about their bots as sentient peers, while teenagers treated them far more casually.
- Women led the trend: Female users showed a significantly higher tendency than men to credit AI companions with human-like emotional depth.
- Joy fuels the illusion: Users who expressed positive emotions—specifically joy—were far more likely to treat the chatbot like a living entity.
- Neutrality acts as armor: Matter-of-fact, emotionally flat interactions strongly correlated with recognizing the bot as a tool rather than a companion.
Teenagers grew up immersed in algorithmic feeds, glitchy bots, and automated feeds; they generally view AI as an interactive search engine or a clever text toy. Adults, by contrast, spent decades associating conversational fluency with human consciousness. When an adult experiences a machine that seems to understand them, they reflexively project a soul onto the code.
This study is an arXiv preprint that has not yet undergone formal peer review. The data comes from public posts in three AI companion subreddits on Reddit, meaning the sample represents people already inclined to use chatbots rather than the general population. Age and gender were inferred or self-reported, which leaves room for demographic error.
- If you catch yourself thanking, flattering, or venting to a voice assistant or chatbot… Switch to concise, functional commands to model for your kids that software is an appliance, not a social connection.
- If your teen uses an AI chatbot for schoolwork or brainstorming… Do not assume they are bonding with it; ask how they use it, and expect to hear that they view it as an imperfect, fast calculator for text.
- If anyone in your home starts speaking about an AI companion with affection or joy… Intervene with factual discussions about how predictive text models work, because positive emotional sentiment is the primary trigger that turns software into a perceived person.
Stop worrying that your teenager cannot tell the difference between code and companionship; grown-ups are actually more vulnerable to the illusion. Treat AI tools in your house like kitchen appliances—useful, functional, and devoid of human warmth—and the whole family will maintain healthy boundaries.
Afia Mubashir, Boden Moraski, Stephanie Choi et al. (2026). Anthropomorphism in AI Companion Communities: Age, Gender, and Emotional Correlates. arXiv (preprint). — arxiv.org