Prince Harry Says AI Chatbots Want Children's Attachment as Expert Warns Kids Are 'Especially Vulnerable' to Mistaking Simulation for Human Connection

Prince Harry is warning parents that artificial intelligence chatbots designed for companionship may present an even greater danger to young people than social media.

“The technology is more persuasive, more intimate, and more powerful,” Harry said September 23 at the Clinton Global Initiative. “AI is learning how to enter our relationships. It doesn’t just want your attention, it wants your attachment.”

The remarks extend Harry’s work through The Parents Network, an Archewell Foundation initiative he launched with his wife, Meghan, in 2024 after speaking about social media’s role in youth anxiety, depression and isolation.

“AI companions can be uniquely risky for kids because they’re always available, highly responsive, and often designed to be agreeable,” said Devorah Heitner, PhD, author of Screenwise and Growing Up in Public.

“Unlike human relationships, they don’t push back, challenge assumptions, or require accountability,” Heitner explained. “That can make children especially vulnerable to mistaking simulated connection for a real relationship.”

She suggested that parents discuss those differences directly rather than simply warning children away from the technology.

“We need to talk with kids and ask questions like: ‘Would you actually want a friend who agreed with everything you said?’ [and] ‘Would you respect someone who never challenged your thinking?’” she advised.

“Parents should watch for kids turning to AI instead of trusted people or their own self soothing strategies — reading, music, exercise, etc. — when they’re lonely, anxious, or upset, especially late at night,” Heitner said.

The concern is not merely how long a child uses a chatbot, but whether the technology begins replacing human support or independent emotional regulation.

Harry said families and advocates have made progress in forcing institutions to recognize technology’s potential harms, but new developments have made their work an “uphill battle.”

“There are major signals that families, young people, experts and advocates have finally shaken institutions awake,” he said. “But bans, settlements and payouts are incomplete remedies.”

Heitner recommended establishing limits on chatbot access while clearly explaining what the technology cannot provide.

“Families should set boundaries around access and emphasize that AI can’t replace real friendships, parents, teachers, or mental-health professionals,” she said. “Children need opportunities to practice empathy, disagreement, and repairing relationships with actual people.”