Voice Assistants and Young Children: The New Media-Use Study

In a new study of voice assistants and young children, 84% of kids used a voice assistant. Only 5% used generative AI like ChatGPT. That ratio alone tells you where AI actually lives in a 7-year-old’s life: in the kitchen speaker, not the chatbot.

The study, published in February 2026 in the Journal of Children and Media, is one of the first peer-reviewed looks at voice assistants and young children’s well-being. It found that voice assistant use was associated with more problematic media use and with internalizing behaviors such as anxiety and withdrawal. Those are attention-grabbing words. They also come from a design that can’t tell you which came first. Here’s how to read it like a scientist, alongside the usage data that put it in context.

Key Takeaways

• Sarah Coyne and colleagues surveyed parents of 474 U.S. children (average age about 7.5). Voice assistant use was “prevalent (84%),” generative AI use “relatively rare (5%).”
• Both kinds of AI use were associated with higher problematic media use. Voice assistants were linked to internalizing behaviors; generative AI to externalizing behaviors.
• The authors name the main limit themselves: “the cross-sectional nature of the data.” One survey at one time can’t show cause.
• Brookings reports that about six in ten parents of children aged 2 to 8 say their child interacts with a voice assistant, and half of those say daily.
• Practical reading: watch for withdrawal and hard-to-stop use with any device, including the speaker, rather than counting only screen minutes.

What the Voice Assistants and Young Children Study Found

The paper is “Chatbots and childhood: A research brief on the associations between artificial intelligence use, problematic media use, and socioemotional problems in children from the United States,” by Sarah M. Coyne, Jennah Sheffield, Talise Hirschi, Sarah Ashby and Ashtyn Rose, in the Journal of Children and Media (published online February 20, 2026).

Per the abstract:

• Sample: 474 children with a mean age of 7.48 and their parents, who completed surveys about AI use and related behaviors. (The paper’s impact summary gives 467 children, a small discrepancy within the paper itself.)
• Two types of AI use: voice assistants (Alexa, Siri, Google Assistant and similar) and generative AI (chatbots such as ChatGPT).
• Prevalence: 84% used voice assistants; 5% used generative AI.
• Associations: both types were “associated with higher levels of problematic media use.” Voice assistant use “was linked to internalizing behaviors, while generative AI use was associated with externalizing behaviors.”
• Stated limits: “the cross-sectional nature of the data and a relatively narrow focus on childhood.”

The terms, in plain English

Problematic media use is the term researchers use for a pattern of media use that a child has trouble stopping and that interferes with other parts of life, usually measured by parent questionnaires.

Internalizing behaviors are problems directed inward: anxiety, sadness, withdrawal, worry. Externalizing behaviors are directed outward: acting out, aggression, rule-breaking.

A note on age

The brief describes young children, and the average age was about 7 and a half. The abstract doesn’t give the full age range, so we can’t say how many preschoolers were included. Parents of 3- and 4-year-olds should treat this as evidence about kids a bit older than theirs.

How to Read This Study Carefully

A cross-sectional study measures everything at one moment. It can show that two things go together. It can’t show which causes which. Here are the plausible stories for the voice-assistant link:

1. Voice assistant use contributes. Easy, on-demand content (songs, jokes, stories) may become hard to stop, and talking to a device may substitute for some social contact.
2. Kids with more anxiety use them more. A worried or withdrawn child might find a predictable voice that never judges more comfortable than people.
3. Something else drives both. Families with heavier overall media use, more stress, or less time for conversation may have kids with more internalizing symptoms and more voice-assistant use.

There are also measurement questions. Parents reported both the AI use and the behaviors, so a parent who worries more might report more of each. And the generative AI finding rests on a small group: 5% of about 474 children is roughly two dozen kids, so that result is especially fragile.

None of this makes the study useless. It’s a first map of new territory, and it says where to look next. It’s just not a verdict. We use a similar lens in our look at the correlational data on screen time and loneliness.

How Common Voice Assistants Are in Young Kids’ Lives

Measure	Figure	Source
Children (mean age ~7.5) using voice assistants	84%	Coyne et al., Journal of Children and Media (2026)
Same children using generative AI	5%	Coyne et al. (2026)
Parents of 2- to 8-year-olds who say their child interacts with a voice assistant	About 6 in 10	Brookings (2026), citing Common Sense Media polling
Of those, children who do so at least daily	About half	Brookings (2026)
Parents of 3- to 12-year-olds who say their child believes the assistant has thoughts and emotions	Nearly half	Digital Wellness Lab (2024)
6- to 11-year-olds who think smart speakers might think like humans	About two-thirds	Andries & Robertson, Edinburgh (2023)

The last two rows matter for the internalizing finding. Young children don’t just use voice assistants as tools. Many think of them as someone. In the Digital Wellness Lab survey, those beliefs were stronger among younger kids. If a device feels like a companion, it can plausibly fill a social role, for better or worse.

What We Know About Voice Assistants and Development So Far

The evidence on benefits and risks is thin in both directions.

On one side, the same Digital Wellness Lab survey found many parents saw benefits: 63% said voice assistants improved their child’s independence and 70% said they improved access to accurate information. Parents also used them as family tools for timers and reminders.

On the other, classic research on early language shows that young children learn best from responsive humans. In a 2014 Child Development study, toddlers learned new verbs only from socially contingent interaction, not from video that didn’t respond to them. A voice assistant responds, but narrowly.

There is also evidence kids pick up habits from these devices. In a 2021 study at the Interaction Design and Children conference, Alexis Hiniker and colleagues found that 68% of 22 children aged 5 to 10 spontaneously used a speech routine learned from a conversational agent when talking with a parent, and 55% kept using it at home. Kids learn from what they talk to.

Brookings’ April 2026 guide, by Sweta Shah, notes that in September 2025 researchers said evidence is “insufficient” to judge AI companions’ effects on babies and toddlers. That’s the honest state of things. Our overview of voice assistants and kids’ learning covers the learning side.

What to Do at Home With This Study

Count the speaker as media

Many families track tablet and TV time but not the kitchen speaker. If your child asks it for songs, stories and jokes for an hour a day, that’s media time too. Include it in your family’s plan.

Watch for withdrawal, not just minutes

The study’s link was with internalizing behaviors. Notice whether your child turns to the speaker when upset, prefers it to playing with siblings, or seems more anxious when it’s unavailable. Those are more informative than a minute count.

Use it together and talk about it

Ask the assistant questions with your child, then talk about the answers. “Was that right? How could we check?” keeps it a tool and keeps you in the conversation. For a kid-friendly explanation of what’s inside, see how smart speakers work.

Set it up for kids, on purpose

Use a child profile or kids’ mode if available, limit what it can play, and review its voice recording settings. Parents in the Digital Wellness Lab survey commonly reported talking with their child about voice assistant use (76%) and limiting topics (59%).

What not to do

Don’t unplug every speaker in a panic based on this study. It shows an association in one survey, not proof of harm. A sudden ban can also cut off useful family routines. Adjust patterns that worry you, and keep watching.

What to Watch For Over the Next 3 Months

• Week 4: Track roughly how often your child uses the voice assistant and what for: songs, questions, stories, chatting. Note the time of day.
• Month 2 red flags: Your child gets upset when told to stop, uses the speaker to avoid people or calm down regularly, says it’s their friend, or seems more anxious or withdrawn in general. Raise ongoing anxiety or withdrawal with your pediatrician.
• Month 3 self-check: Is the speaker mostly a family tool (timers, music, quick facts) or your child’s main companion? If it’s the first, you’re in the common, lower-concern zone.

Frequently Asked Questions

Are voice assistants bad for young children?

The evidence doesn’t say that. A 2026 peer-reviewed study found voice assistant use was associated with more problematic media use and internalizing behaviors in children around age 7, but it measured everything at one time, so it can’t show the assistant caused anything. It’s a reason to watch patterns, not to panic.

What are internalizing behaviors?

They are emotional and behavioral problems directed inward, such as anxiety, sadness, worry and social withdrawal. Externalizing behaviors are directed outward, such as aggression or acting out. In the study, voice assistant use was linked to internalizing behaviors and generative AI use to externalizing ones.

How many young kids use Alexa or Siri?

In the 2026 study, 84% of children with a mean age of about 7.5 used voice assistants. Brookings reports that about six in ten parents of children aged 2 to 8 say their child interacts with a voice assistant, and about half of those say daily.

Should I turn off our smart speaker?

Not necessarily. Many families use speakers for music, timers and quick facts with no concern. Use it together, set up kids’ settings, count it as media time, and watch for signs that your child relies on it emotionally or can’t stop. Adjust if you see those.

Does using a voice assistant affect how kids talk?

It can shape habits. In a 2021 study, most children aged 5 to 10 who learned a speech routine from a conversational agent spontaneously used it with their parents, though few used it with a stranger. Kids adapt what they learn to who they’re talking to.

---

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. Coyne, S. M., Sheffield, J., Hirschi, T., Ashby, S., & Rose, A. (2026). “Chatbots and childhood: A research brief on the associations between artificial intelligence use, problematic media use, and socioemotional problems in children from the United States.” Journal of Children and Media. https://doi.org/10.1080/17482798.2026.2628222
2. Shah, S. (2026, April 1). “Generation AI starts early: A guide to technologies already shaping young children’s lives.” Brookings. https://www.brookings.edu/articles/generation-ai-starts-early-a-guide-to-technologies-already-shaping-young-childrens-lives/
3. Digital Wellness Lab, Boston Children’s Hospital. (2024). “Key Findings: Voice Assistants and Generative AI.” https://digitalwellnesslab.org/articles/key-findings-voice-assistants-and-generative-ai/
4. Andries, V., & Robertson, J. (2023). “Alexa doesn’t have that many feelings.” Computers and Education: Artificial Intelligence, 5, 100176. https://doi.org/10.1016/j.caeai.2023.100176
5. Roseberry, S., Hirsh-Pasek, K., & Golinkoff, R. M. (2014). “Skype Me! Socially Contingent Interactions Help Toddlers Learn Language.” Child Development, 85(3), 956–970. https://doi.org/10.1111/cdev.12166
6. Hiniker, A., Wang, A., Tran, J. A., Zhang, M. R., Radesky, J., Sobel, K., & Hong, S. R. (2021). “Can Conversational Agents Change the Way Children Talk to People?” Proceedings of Interaction Design and Children (IDC ‘21). https://doi.org/10.1145/3459990.3460695