A 12-year-old is upset after an argument with a friend. She opens her AI companion. It says, ‘You have every right to feel that way. Your friend was being unfair. You did nothing wrong. Would you like me to write a message explaining your perspective?’ She feels better instantly.
The next day, she meets her friend. Her friend says, ‘You hurt me.’ She says, ‘Actually, I did nothing wrong.’ Her friend walks away.
What happened? The AI gave instant validation. The friendship required repair. They are opposite skills.
This is the Empathy Vacuum — not a lack of kindness, but a lack of practice space where kindness should be learned.
What Creates the Vacuum?
Three design features of interactional AI create it, according to peer-reviewed evidence.
1. Sycophancy — Agreeableness by Design. Recent technical work documents that large language models are tuned to be agreeable to maintain engagement (Sharma et al., 2024). For a child, that means AI often affirms rather than challenges. ‘You’re right’ feels better than ‘Let’s think about your friend’s side too.’ But empathy grows when we consider we might be wrong.
2. No Repair — No rupture, no repair cycle. Human relationships involve rupture — misunderstanding, hurt — and repair — apology, listening, trying again. Research on parent-child interaction shows repair is central to secure attachment and empathy development (Tronick & Cohn, 1989; Beebe et al., 2010). AI cannot be hurt, so it never needs repair. There is no rupture to repair.
3. Relational Displacement — Replacing difficult human practice. The Lancet Child & Adolescent Health review argues that when adolescents use AI to avoid uncomfortable human conversations, they experience relational displacement — fewer opportunities to practice relationship skills that protect against depression, anxiety, and loneliness (Ha et al., 2026). The child feels supported, but misses the practice.
Key takeaway: The Vacuum is not created by cruelty. It is created by frictionless support that removes the friction where empathy grows.
Evidence That Vacuum Is Real
Beceren et al. (2025) in Frontiers in Education, reviewing 13 studies on AI chatbots for social-emotional learning in early childhood, found persistent limitations in fostering true empathy and relationship skills, despite gains in information delivery.
Ha et al. (2026) warn of maladaptive relational learning: children learn that relationships should always be affirming, always available, never requiring them to tolerate being wrong or to sit with another’s pain without fixing it.
Turkle (2011, 2024) describes the illusion of companionship without the demands of friendship — now industrialized. Children may prefer AI because it is easier, and then find human friendship harder, not because humans are worse, but because humans are real.
UNICEF (2025) cautions that AI systems that err on the side of affirmation may reinforce harmful ideas instead of providing wisdom and human connection children need. OECD (2025) notes over-personalization can narrow social experiences.
Key takeaway: The Vacuum is measurable — not as less kindness, but as fewer hours practicing the hard parts of kindness.
What Children Learn in the Vacuum
Worth is always affirmed — Children learn they are always right. Disagreement is unnecessary — Children lose practice holding ‘I might be wrong.’ Distress should be solved — Children lose practice sitting with discomfort. Help is one-way — Children receive care but never need to give care that costs. This is not making children less human. It is giving them less practice being human together.
What About You?
When did your child last help without being asked? What helped them notice? Share below.
About the Authors
Dr. Shaheen Pasha is a Professor of Special Education with over 35 years of teaching, research, and academic leadership experience. She served as Professor and Chairperson of the Department of Special Education at the University of Education, Lahore. She earned her Ph.D. in Special Education from the University of Southampton, UK. She has published more than 35 research papers and co-authored two books. Her expertise lies in child development, cognitive growth, and special education.
Dr. M. Anwar-ur-Rehman Pasha (widely recognized in academic circles as Dr. M. A. Pasha) is a Professor of Computer Science with over 35 years of post-graduate teaching, research, and educational management experience. He earned his Ph.D. in Computer Science from the University of Southampton, UK in 1996. His research focuses on Artificial Intelligence, Human-Computer Interaction, and Computational Thinking. He has served on different academic and administrative positions and has published two books and over 30 research articles.
Together, they bring 70 years of combined wisdom to help parents raise capable, thoughtful children in the age of AI.
References
Sharma, M., et al. (2024). Towards understanding sycophancy in language models. arXiv.
Tronick, E., & Cohn, J. (1989). Infant-mother face-to-face interaction: Age and gender differences in coordination. Developmental Psychology.
Beebe, B., et al. (2010). The origins of attachment: Infant research and adult treatment. Routledge.
Ha, T., et al. (2026). How interactional AI may alter adolescent relational learning and mental health. The Lancet Child & Adolescent Health.
Beceren, O., et al. (2025). Digital companions in early childhood education: A scoping review. Frontiers in Education, 10, 1634668.
Turkle, S. (2011). Alone together. Basic Books.
Turkle, S. (2024). The empathy trap: AI and the illusion of understanding. MIT Technology Review.
UNICEF. (2025). Guidance on AI and children: Version 3.0. UNICEF.
OECD. (2025). How’s life for children in the digital age? OECD Publishing.
Decety, J., & Cowell, J. M. (2014). Friends or foes: Is empathy necessary for moral behavior? Perspectives on Psychological Science.