玩具盒中的人工智能：父母如何看待面向幼儿的人工智能赋能玩具

AI in the Toy Box:How Parents View AI-Enabled Toys Acknowledgements Common Sense Media is grateful for the generous support and underwriting that funded this research spotlight: Bezos Family FoundationCraig Newmark PhilanthropiesJennifer Caldwell and John H.N. FisherPatrick J. McGovern FoundationSiegel Family Endowment Credits Data collection: SSRS Designers: Mina Cheong & Chris Arth Copy editor: Christopher Dare Communications and Marketing: Christian Medina Beltz, Kirsten Olsen Cahill, Lisa Cohen, Edda Collins Coleman,Erika Radhansson, Brian Ray, and Owen Stidman Suggested citation: Robb, M. B., & Mann, S. (2026).AI in the toy box: How parents view AI-enabledtoys for young children.San Francisco, CA: Common Sense Media. Introduction AI-enabled toys and child-focused devices are quickly moving from novelty to mainstream holiday offerings,marketed as playful, educational, and increasingly conversational companions for young children. Recent mediacoverage has highlighted both excitement and discomfort with this trend, as parents encounter toys that talkback, remember past interactions, and simulate social or emotional responsiveness (Hess & Messinger, 2025;Radesky, 2025; Suskind, 2025). At the same time, early safety testing has begun to surface potential risks specific National research on teensfinds that a large majority have already interacted with AI companions, often usingthem for curiosity-driven exploration, advice, or emotional support (Robb & Mann, 2025). These patterns of useraise important questions about emotional dependency and social development. Younger children may face evengreater risks: Developmental research suggests that younger children may be even more susceptible to Common Sense Media's current AI risk assessments recommend that AI companions not be used by anyoneunder 18, reflecting concerns about emotional dependency, unsafe content, and data practices (Common SenseMedia, 2025). As AI becomes embedded not only in apps but also in physical toys and play environments, it isimportant to examine whether parents perceive AI-enabled toys differently from AI companions, and under what This report draws from a nationally representative survey of 1,004 parents, conducted December 1 to 8,2025, to explore the perspectives of parents of children age 0 to 8 on AI-enabled toys. Specifically, itexamines their current exposure to these toys, their purchase intentions, the hopes and concerns they hold,and the trade-offs they are currently weighing. The survey was conducted via web in English (n= 951) and Respondents received the following definition: "When we say 'AI-enabled toys or devices,' we are NOTreferring to general-purpose voice assistants like Alexa or Siri. We are asking specifically about toys orchild-focused devices designed for kids, like 'smart' toys, robots, stuffed animals, or other physical toys Key Findings 1.Nearly half of parents (49%) have purchased or are considering AI-enabled toys/devices A similar proportion (43%) say they do not own and do not intend to buy AI-enabled toys. Parents of olderchildren (age 5 to 8) are more likely than those of younger children (age 0 to 4) to already own AI-enabled toys(19% vs. 12%). Black parents report the highest levels of adoption and purchase intent (78%), compared toHispanic (61%) and White parents (36%). Future interest is higher among lower-income households (those that Table 1: AI toy ownership or purchase intent Note: Columns tested a/b/c – d/e/f – g/h. Items with different superscripts differ significantly (p < .05). Significance should be across rows, within demographic groups(race/ethnicity, household income, child age). Total amounts may not sum to 100% due to rounding or nonresponse. 2.Nearly one infive (19%) parents say they would at least somewhat like an AItoy/device to be a companion for their child. More than half (56%) of all parents say they do not want AI-enabled toys/devices to serve as a companion orsidekick for their child. Parents whose children already use AI toys daily are more open to companionshipfunctions: 48% express at least some interest, compared to 23% of weekly or less frequent users and 12% ofnon-users. Resistance is higher among parents of younger children, with 84% of parents of 0- to 4-year-olds Table 2: Parents who want AI-enabled toys to serve as a companion for their child Note: Columns tested a/b/c – d/e/f – g/h. Items with different superscripts differ significantly (p < .05). Significance should be across rows, within demographic groups(race/ethnicity, household income, child age). Total amounts may not sum to 100% due to rounding or nonresponse. An asterisk (*) indicates greater than 0 but less 3.Collection of personal information is parents' top concern about AI toys, with 83% Other top concerns are challenges with setting usage limits (79% concerned) and cybersecurity risks (79%). Thepossibility of toys/devices saying inappropriate things to their child concerns 74% of