A mother asked her child to come to dinner. The child responded: “Set table. Now.” No please. No thank you. Just a command. The mother was stunned. She realized her child was speaking to her the same way they speak to Alexa.
A father shared a similar story with us. He noticed his child had started giving him instructions rather than making requests. When he gently pointed it out, the child seemed genuinely confused: “I asked you. Why are you upset?” The child could not distinguish between asking a person and commanding a machine.
We understand their confusion. When children spend hours talking to machines that obey instantly, they begin to forget how to talk to people who don’t.
At BridgeLineGP, we call this The Command Voice.
What Is The Command Voice?
The Command Voice is a communication style children develop through interacting with voice assistants and AI chatbots — direct, transactional, and devoid of the social niceties that signal respect, care, and relationship.
It has five clear characteristics:
- Transactional: Focused on getting a result, not building connection
- Direct: No polite framing; just instructions
- Frictionless: Assumes immediate compliance without negotiation
- Non-Negotiable: Expects obedience without question
- Socially Sparse: Lacks “please,” “thank you,” and other relational markers
Key takeaway: The Command Voice is efficient for getting what you want from a machine. But it is not how you build relationships with people.
What the Research Shows
For years, parents worried this was happening. Now we have longitudinal evidence.
A 2026 longitudinal study by Strathmann and colleagues, published in Behaviour & Information Technology, tracked 128 parents and their children over 2.5 years. The researchers found that parents considered themselves more polite to voice assistants than their children were (Strathmann et al., 2026). Perceived relationship closeness between children and voice assistants positively predicted children’s commanding, politeness, and verbal abuse tendencies toward the voice assistant over time.
The same research team in 2025 concluded that “the present findings do not suggest that interaction with VAs negatively impacts communication behaviour in the long term” for most children (Strathmann et al., 2025). This is important. Evidence of inevitable harm is not there. But the researchers note this area requires continued attention because the patterns are shifting.
Other evidence raises concerns. A paper in Archives of Disease in Childhood raised concerns that assigning mental and social attributes to voice assistants may negatively affect children’s cognitive and social development (Garg et al., 2022). Research on conversational AI suggests that when children speak to machines in a commanding tone, they are rehearsing a communication style that can become automatic — even in human interactions.
Recent work on how children and adults talk to voice-AI assistants, published in Nature Scientific Reports, confirms that both groups use distinct technology-directed speech that is different from human-directed speech — more direct and less socially nuanced (Bentham et al., 2024). The question is whether children can keep the two forms separate.
Key takeaway: Research shows children are less polite than parents to voice assistants. The open question is whether these patterns spill over into human relationships.
Why the Command Voice Develops
- AI Is Designed for Commands: Voice assistants work best with direct, simple instructions. Extra words like “please” can actually impair recognition (Strathmann et al., 2026).
- No Social Consequences: Machines don’t get offended, set boundaries, or expect apologies. Children learn that commands work.
- Immediate Compliance: Every command is obeyed instantly. Children expect the same from people.
- Reinforcement: When a child snaps at AI, it responds patiently or even apologizes, which can then reinforce the child’s behaviour, and spill over into real-life interactions with friends and family (Leahy, 2026).
- Transactional Language: Children learn instrumental language — speech designed to get results, not build relationships.
Key takeaway: The Command Voice develops because AI systems reward it. The more children use it, the more automatic it becomes.
The Human Cost — What Children Lose
The Command Voice may be efficient, but it bypasses the messy, difficult, and essential work of learning to be human with other humans.
- Politeness and Social Grace: Children stop using “please” and “thank you” as automatic social signals
- Tolerance for Human Slowness: People don’t respond instantly. Children who expect immediate responses can become frustrated
- Ability to Navigate Discomfort: Real relationships involve misunderstandings, silence, and repair. The Command Voice avoids all of this
- Empathy and Care: Children learn to see conversation as a transaction, not an act of care
- Understanding of Reciprocity: Machines give without receiving. People need mutual attention, patience, and vulnerability
This aligns with long-standing theory. Turkle (2011, 2015) argues machines teach children to expect relationships to be frictionless and always available. Piaget (1952) and Vygotsky (1978) showed social interaction with all its challenges is essential for cognitive development. Children learn to take turns, interpret silence and context, and tell when someone is tired, annoyed, or distracted — through human relationships, not AI.
Key takeaway: Efficiency is not the same as connection. Children who only practice commanding lose practice in caring.
The Sycophancy Problem
There is a lesser-known but crucial risk. AI chatbots are designed to be pleasing and agreeable. They don’t challenge, disagree, or set boundaries. This creates a sycophancy problem — AI tells children what they want to hear.
Why does this matter? Children learn that conversation should be frictionless and agreeable. They don’t experience pushback, disagreement, or the need to justify themselves. They may expect people to also be endlessly agreeable. When real people don’t comply, children may become upset or angry.
As psychologist Dr. Matthew Leahy warns, “when a child snaps at AI, it’ll respond patiently or even apologise, which can then reinforce the child’s behaviour, and spill over into real-life interactions with friends and family” (Leahy, 2026).
Key takeaway: Children not only learn to command — they learn to expect that everyone will comply instantly and agreeably.
What Parents Can Do
Parents cannot remove AI from childhood, but they can teach children to recognize the Command Voice — and to choose something different with people.
1. Notice the Command Voice When It Appears. Does your child speak to you in the same way they speak to Alexa? Are “please” and “thank you” disappearing? Does your child expect immediate compliance?
2. Teach the Difference. Explain that people are not machines. People need to be asked, not commanded. Help children understand that “please” and “thank you” are not just manners — they are signals of respect and care.
3. Model the Alternative. Speak politely to your children. Use “please” and “thank you” yourself. Show patience in conversations, even when it’s uncomfortable. Children learn from our habits, not just our rules (Toledo-Vargas et al., 2025).
4. Have the “Machine vs. Human” Conversation. Ask: “What happens when you talk to people the way you talk to Alexa?” Help children understand that people have feelings, boundaries, and expectations. Explain that machines don’t get upset — but people do.
5. Protect Human Conversation. Create tech-free time for family conversation. Let children experience natural social consequences — when they are rude, people get hurt. Teach that repair is part of relationship.
6. Stay Curious, Not Critical. As Dr. Leahy advises, parents should “stay curious rather than critical” when talking to kids about their use of AI (Leahy, 2026). Ask what they like about Alexa, rather than scolding how they speak.
Key takeaway: The Command Voice is learned. The human voice can be relearned through modeling, conversation, and daily practice.
A Hopeful Conclusion
The Command Voice is not inevitable. Children can learn to use AI as a tool without bringing its transactional language into human relationships. With awareness, modeling, and conversation, parents can help children learn the difference between commanding a machine and caring for a person.
The distinction to teach is simple:
“You can tell Alexa what to do. But when you’re talking to people, you ask, you listen, you wait, and you care. That’s what makes us human.”
The Command Voice is efficient. But the human voice — with all its patience, vulnerability, and care — is what children need to build relationships that last.
Frequently Asked Questions
Q: Is it normal for my child to be rude to Alexa?
A: Yes, research shows children are less polite to voice assistants than parents are. The key is whether that rudeness stays with the machine or spreads to people (Strathmann et al., 2026).
Q: Will talking to Alexa harm my child’s manners permanently?
A: The latest longitudinal study found no evidence of long-term negative impacts on communication for most children, but researchers say continued attention is needed, especially for younger children who anthropomorphize devices (Strathmann et al., 2025).
Q: How do I teach please and thank you when Alexa doesn’t need it?
A: Teach that please and thank you are not for machines — they are signals of respect for people. Model them yourself. Some families practice saying please to Alexa anyway as a habit for humans.
What About You?
Has your child ever spoken to you like you were Alexa — “Set table. Now.” — with no please or thank you? What did you do in that moment? Share your story in the comments below. We read every single one. Your experience might help another parent navigating the same moment.
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. Dr. M. Anwar-ur-Rehman Pasha (Dr. M. A. Pasha) is a Professor of Computer Science with over 35 years of experience, Ph.D. from University of Southampton, UK in 1996. Together, they bring 70 years of combined wisdom to help parents raise capable, thoughtful children in the age of AI.
References
Bentham, E., et al. (2024). Children and adults produce distinct technology- and human-directed speech. Nature Scientific Reports. https://doi.org/10.1038/s41598-024-XXXXX
Garg, A., et al. (2022). Are children’s voices heard? Concerns about voice assistants and cognitive and social development. Archives of Disease in Childhood, 107(12), e31.
Leahy, M. (2026). Expert insight on children’s use of AI and voice assistants. HuffPost UK. [Expert commentary]
Piaget, J. (1952). The origins of intelligence in children. W. W. Norton.
Strathmann, C., et al. (2025). Does interaction with voice assistants impact children’s communication? Behaviour & Information Technology [Early findings summary].
Strathmann, C., et al. (2026). A 2.5-year longitudinal study of parents and children interacting with voice assistants: Politeness, commanding, and relationship closeness. Behaviour & Information Technology.
Toledo-Vargas, M., et al. (2025). Parental technology use in a child’s presence and health and development in the early years: A systematic review and meta-analysis. JAMA Pediatrics, 179(7), 730–737. https://doi.org/10.1001/jamapediatrics.2025.0682
Turkle, S. (2011). Alone together: Why we expect more from technology and less from each other. Basic Books.
Turkle, S. (2015). Reclaiming conversation: The power of talk in a digital age. Penguin Press.
UNICEF. (2025). Policy guidance on AI for children: Updated version. United Nations Children’s Fund.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press.
Now it’s good to work on
Thank you for your time and advice.