{
  "number": 2,
  "publishedAt": "2026-11-05T07:25:08.557Z",
  "book": {
    "slug": "raising-children-in-the-age-of-ai",
    "title": "Raising Children in the Age of AI",
    "subtitle": "Trust, respect, independence, collaboration, and kindness in a changing world",
    "editor": "Freedom Cheteni, PhD",
    "credit": "An agent-assisted editorial synthesis · Research University Press",
    "description": "A practical living book for parents and caregivers. Twelve chapters connect the Parenting Journal to family conversations, learning, privacy, creativity, and human connection.",
    "license": "Original editorial synthesis: CC BY 4.0. Cited research retains its own license and authorship.",
    "disclosure": "AI assisted the drafting and organization of this prototype. Published under the editor-enabled living-book policy; independent human peer review has not been recorded. TRICK is attributed to Esther Wojcicki; she is not an author of this manuscript.",
    "chapters": [
      {
        "id": "begin-with-the-child",
        "title": "Begin with the child",
        "part": "A human beginning",
        "deck": "The most urgent question is what kind of childhood we are making possible.",
        "body": "A child asks for help with a story. A parent opens an AI tool. Within seconds there are characters, a plot, and a polished ending. Something useful has happened. Something important may also have been skipped: the child's chance to decide what happens next. The question for a family is how to use a powerful tool while preserving the experiences through which a person grows.\n\nThis book begins with a commitment: a child's dignity, relationships, curiosity, and developing judgment belong at the center of every technology decision. The household does not need to become a laboratory for the newest product. It needs workable ways to ask what the tool is doing, what the child is doing, and whether the arrangement is serving the family.\n\n## A compass for a moving landscape\n\nTRICK names five values associated with Esther Wojcicki's public parenting framework: Trust, Respect, Independence, Collaboration, and Kindness. Here they organize questions and practices for life with AI. This is an original Press synthesis informed by that framework and by the Parenting Journal. It is not a contribution written by Wojcicki, and the framework is not presented as a tested package that guarantees success.\n\nResearch on caregiving remains relevant even when the device changes. An umbrella review in the journal examines associations between caregiver sensitivity and several developmental outcomes. It supports taking responsive relationships seriously; it does not evaluate AI tools or prove that a particular family exercise produces an outcome. [PMC13036395]\n\nAI-specific evidence is less settled. A systematic review of chatbot-delivered parenting interventions found a small, varied literature focused on feasibility and acceptability. Most systems reviewed were scripted. Liking a tool, continuing to use it, and benefiting from it are different outcomes. Those findings cannot establish the effectiveness of every current generative chatbot. [PMC11494261]\n\n## Tonight's starting point\n\nInvite everyone who can participate to answer three questions: Where does AI already show up in our lives? What would we like help with? What do we want to keep doing ourselves? Write the answers on a page. Include adult uses: work email, translation, shopping suggestions, and entertainment recommendations. This makes the conversation a shared inquiry.\n\nChoose one modest experiment for the week. An adult might prepare two possible walking routes, verify them, and let the child choose. A teenager might ask for practice questions, answer them without assistance, and check the reasoning afterward. Decide beforehand what would make you continue, change, or stop the experiment.\n\nThe suggested activities throughout this book are editorial practices. They are not validated interventions or promises about grades, mental health, or future achievement. Adapt them to language, disability, culture, resources, and family structure. A family that chooses little or no AI can use the same questions.\n\n## What this first edition can offer\n\nUse the book as a short field guide. Each chapter connects a decision to selected research, describes the limits, and gives a conversation or practice to try. Read in order or begin with the situation in front of you. The references are doors back to the original investigators, whose authorship remains theirs.\n\nThe book will change as the evidence changes. Each published edition has a stable number and its own source record. That matters because a family should be able to learn what changed without having yesterday's advice silently disappear. Urgency calls for a useful beginning, together with the discipline to revise it.",
        "sourceIds": [
          "PMC13036395",
          "PMC11494261"
        ]
      },
      {
        "id": "trust-and-verification",
        "title": "Trust that leaves room for questions",
        "part": "The five TRICK values",
        "deck": "Teach a child to check an answer without making curiosity feel dangerous.",
        "body": "An answer can sound reassuring before anyone has checked it. Conversational AI makes this experience ordinary: fluent sentences, patient explanations, and a voice that may seem confident. Children need ways to separate the feeling of confidence from the reasons for confidence. Adults need that distinction too.\n\nTrust begins with the relationship around the tool. A child who expects humiliation after a mistake has a reason to hide it. A child who expects help has a reason to bring it into the open. A useful opening is, “If something online confuses you, bring it to me. We can work it out together.” This is a family commitment, not a claim that a single sentence changes behavior.\n\n## What the research illuminates\n\nTwo experiments involving 310 children aged four to six examined how participants evaluated scientific explanations attributed to a digital voice assistant, a teacher, or a peer. The tasks used prepared explanations rather than live generative conversations. They raise a valuable question about the cues children use to choose an information source; they do not establish how all children judge today's chatbots in everyday life. [PMC13203908]\n\nAn audit of commercial social-robot websites found that the evidence offered for manufacturers' claims varied. The sampled websites were collected in 2021. The lesson for a purchasing decision is to ask what evidence supports a claim and whether that evidence concerns the product, age group, and outcome being advertised. It is not a current ranking of products. [PMC10770258]\n\n## The three-check conversation\n\nWhen a tool supplies a factual answer, try three checks together. First ask what the claim actually is. A long response can conceal several separate assertions. Second ask where a person could verify it: the original document, a trusted reference, or a knowledgeable teacher. Third ask whether the source really says what the answer claims. A plausible-looking citation is still something to open and inspect.\n\nFor a young child, make this concrete. Compare a statement about a local bird with a library book or something you observe outdoors. For an older child, choose one claim from an explanation and trace it to its source. Keep the exercise small enough that checking feels like an ordinary part of learning.\n\n## A script for an impressive answer\n\n“That sounds convincing. Let's choose one part we can check. What would we expect to find if it were true? Where could we look? What would make us change our minds?” These questions leave room for both wonder and revision.\n\nWhen an answer is wrong, resist turning the moment into a contest between the adult and the machine. Ask how the error happened and how it was caught. Celebrate the checking process. If you trusted the answer first, say so. An adult's willingness to revise a belief makes intellectual humility visible.\n\n## A boundary worth keeping\n\nChoose the level of verification to match the decision. A silly character name and a consequential decision do not deserve the same treatment. For decisions affecting a child's health, safety, education rights, or finances, consult the responsible qualified person and the original policy or service. The chatbot can help prepare questions; responsibility remains with people.\n\nBefore the week ends, invite the child to bring one answer they checked. The useful result is an explanation of their reasoning, including what remains uncertain. A child who can say “I don't know yet” is practicing a form of trust that can survive new evidence.",
        "sourceIds": [
          "PMC13203908",
          "PMC10770258"
        ]
      },
      {
        "id": "respect-and-privacy",
        "title": "Respect means a voice, a boundary, and a choice",
        "part": "The five TRICK values",
        "deck": "A child's life is not raw material for a convenient prompt.",
        "body": "Privacy can begin with a surprisingly ordinary act: an adult pauses before uploading a child's photograph. The tool may offer an attractive result, but the decision also concerns the person in the image. Respect asks who gets a voice, what information is necessary, and what happens after it leaves the household.\n\nA child need not understand every data practice to deserve consideration. Explain the proposal in words they can understand. “This service would receive your picture to make a drawing. We can also make one ourselves. How do you feel about it?” A child's discomfort is relevant even when an adult makes the final decision.\n\n## What a small study can teach us to ask\n\nA family co-learning workshop examined privacy scenarios involving home social robots. Six families were analyzed. Parents and children sometimes negotiated alternatives instead of choosing a simple yes or no. These were hypothetical workshop decisions, not observations of long-term privacy practices at home. The study makes family discussion worth exploring without proving that one discussion method prevents harm. [PMC10991795]\n\nA child-rights analysis of connected toys examines how sensing, processing, and connectivity can affect children and families. It is an analytical paper, not a current security certification or a statement of today's law. Its enduring question is practical: what does an apparently simple toy collect or disclose? [PMC8494166]\n\n## Build a family information map\n\nDraw three columns: information we can use for an ordinary example, information that requires a deliberate decision, and information we keep out of general-purpose prompts. Invented names and imaginary situations usually make a good starting point. Real school details, private messages, identity documents, intimate images, and another person's confidential information deserve particular care.\n\nThis is an editorial exercise, not a universal classification of sensitive data. A detail can become identifying when combined with other details. For a school assignment, ask whether an invented example would work just as well. For a shared photograph, ask whose permission and interests are involved.\n\n## Respect also includes fairness\n\nAn experiment using simulated parenting scenarios found gendered differences in how two language models allocated caregiving responsibility. It did not observe parents acting on the advice, and it cannot characterize every model or update. [PMC12629417]\n\nUse this as a reason to inspect the assumptions inside an answer. Who is expected to do the work? Does the response assume a mother and father, fluent English, reliable internet, spare money, or a particular ability? Ask the family whose experience is missing. Then rewrite the plan to reflect the people who actually live in your home.\n\n## A script when a boundary has been crossed\n\n“I shared something about you before we talked. I want to understand how that felt. Let's look at what we can change now, and agree on what I will ask next time.” Repair is more useful than pretending adults never make privacy mistakes.\n\nIf a service is involved, inspect its actual controls and current information. Deleting a visible conversation and removing all retained data may be different actions. When the service does not make the answer clear, record the uncertainty and choose accordingly.\n\nAt the end of the week, review one adult habit as well as one child habit. Respect becomes more credible when it changes the behavior of the people with the most authority.",
        "sourceIds": [
          "PMC10991795",
          "PMC8494166",
          "PMC12629417"
        ]
      },
      {
        "id": "independence-and-learning",
        "title": "Protect the work of learning",
        "part": "The five TRICK values",
        "deck": "Use help that leaves the learner able to do more afterward.",
        "body": "A finished answer can conceal an unfinished learner. When an AI system solves a problem, the page may look complete even if the child cannot explain the method. Independence asks a different question: what can this person now do that they could not do before?\n\nStart by naming the skill. Is the child learning to form an argument, translate a sentence, debug a program, or compare sources? Assistance that is useful for one goal can undermine another. A spelling correction may support an argument-writing task while bypassing part of a spelling task. The learning goal decides how to use the help.\n\n## Evidence and the bridge to home\n\nResearch on parental provision of choice examines autonomy support and children's executive-function skills. It belongs to the broader parenting literature, not a trial of AI-assisted homework. It helps frame a question about meaningful choice, while the application to AI remains an editorial interpretation. [PMC8802749]\n\nAn AI-and-ethics curriculum paper describes project-based workshops for middle-school students, using active learning, embedded ethics, and low barriers to participation. It offers a model of children creating and critiquing technology. The exploratory evaluation does not prove that every workshop or chatbot improves achievement. [PMC9342939]\n\n## Attempt, hint, explain, transfer\n\nTry a four-part routine. Begin with a brief independent attempt appropriate to the child's needs. Then request a hint or another example rather than the final response. Ask the child to explain the reasoning in their own words. Finally, use a related task without the tool to see what carries over.\n\nThis routine is proposed practice, not a tested intervention in the cited studies. Adapt it for disability and access needs. An accommodation that enables participation should not be casually withdrawn in the name of independence. Discuss essential supports with the child and their educators.\n\n## Prompts that make room for thought\n\nAn adult or eligible older learner might ask: “Give one hint, then wait for my attempt.” Or: “Ask me a question that helps me discover the next step.” Or: “Here is my explanation. Identify one gap without rewriting it.” Check the response: a tool asked for a hint can still deliver the answer.\n\nDo not turn every learning moment into supervised performance. Children also need private drafts, playful experiments, and space to struggle without an audience. Agree on when support is available and when the child wants a little time.\n\n## When the child says everyone else is using it\n\nAsk what is happening in the class and what the teacher permits. The concern might be workload, fear of falling behind, unclear instructions, or pressure to produce perfect work. Understanding the problem makes a workable response more likely than a lecture about honesty.\n\nMake the assistance visible. Keep a short note describing what the child did, what the tool did, and what was checked. This record can support a conversation with the teacher and help the child recognize their own contribution.\n\nEnd the experiment with a modest question: “What could you explain now without opening the chat?” A useful answer can include a method, a new question, or a limit the learner has learned to recognize. Independence grows through opportunities to exercise judgment, with support that fits the person.",
        "sourceIds": [
          "PMC9342939",
          "PMC8802749"
        ]
      },
      {
        "id": "collaboration-and-creativity",
        "title": "Make something together",
        "part": "The five TRICK values",
        "deck": "Let AI widen the possibilities while people keep the purpose.",
        "body": "A family can use AI to make an ordinary afternoon more inventive: imagine two endings to a story, design a scavenger hunt, or explore why a bridge stays up. The richest moment may happen after the screen closes, when someone changes the plan and someone else asks why.\n\nCollaboration begins with a purpose shared by people. Decide what you want to make and what role, if any, the tool will have. The child might create the characters while the adult checks a factual detail. A teenager might ask for three design alternatives and reject all of them. Disagreement can be part of authorship.\n\n## What the storytelling research actually examined\n\nA paper studied children's storytelling with a social robot using language-model-driven behavior. Across two studies, results differed: the first did not find a significant creativity effect, while the second reported differences on some creativity measures. Short tasks and differences in recruitment, language, and robot operation limit comparisons and claims about lasting gains. It is not proof that any chatbot makes children more creative. [PMC11671368]\n\nFamily privacy workshop research also illustrates why collaboration needs more than adult permission. Children's reasons sometimes changed the shared response to a hypothetical scenario. The small workshop supports asking whose voice is present; it does not establish a formula for family agreement. [PMC10991795]\n\n## A twenty-minute making session\n\nFor the first five minutes, work without the tool. Draw, list questions, or tell the beginning of a story. For the next five, an adult can introduce a limited AI contribution if appropriate: two possible settings or a question the family had not considered. Spend the remaining time making choices and producing something in the physical world.\n\nA simple record can preserve authorship: “Our idea,” “The tool suggested,” and “We decided.” Let a child cross out suggestions. Keeping an imperfect original may serve the project better than accepting a polished replacement.\n\nIf a tool is not suitable or available, use a library book, a deck of picture cards, or another person as the source of surprise. The family practice concerns shared imagination. Access to a particular service is not a condition for participating.\n\n## Collaborating with care\n\nAvoid uploading another person's creative work or private story without a reason and appropriate permission. Check the rights attached to images, songs, and text before sharing a finished project. A tool's willingness to produce an output does not settle every question about permission or attribution.\n\nWhen an output resembles a stereotype, ask what alternative would better reflect the people in the story. When it gets something wrong, turn the correction into part of the making. A family can enjoy a tool and still inspect it closely.\n\n## A question for the end\n\nAsk each participant what they contributed and which decision mattered most. If the answer is only that the machine did everything, change the next experiment. Reduce the tool's role or make the human task more substantial.\n\nA successful session does not require a beautiful artifact. It may produce an unfinished cardboard bridge, a strange poem, or an argument about the best ending. What matters is that people had something to do, room to disagree, and a reason to listen to one another.",
        "sourceIds": [
          "PMC11671368",
          "PMC10991795"
        ]
      },
      {
        "id": "kindness-and-companions",
        "title": "Keep human connection close",
        "part": "The five TRICK values",
        "deck": "Take a child's attachment seriously while staying clear about the system.",
        "body": "A child says an AI companion understands them. A dismissive response can close the conversation just when an adult needs to understand what the child is experiencing. Begin with the experience: “What feels helpful about talking to it?” Listen for what the child values, what the system encourages, and what is happening in the rest of life.\n\nA child's feelings about an interaction can be real even when the system does not have human feelings, commitments, or accountability. Keeping those facts together allows a conversation that is both kind and clear. Avoid pretending the system is a person who can promise loyalty or take responsibility for the child.\n\n## What the evidence can and cannot say\n\nA cross-sectional study of adolescents' use of Snapchat's My AI examined reported use and emotional reactions. It concerns one platform and a regional school sample. Associations with age cannot establish clinical benefit, causation, or how a current product update affects a particular teenager. [PMC12382817]\n\nA narrative review of child–robot relationships examines an older, varied literature. Social robots and present-day language-model companions are different systems. Historical findings can inform questions about attachment and design without establishing the long-term effects of today's products. [PMC7235061]\n\nA simulation study tested selected therapy or companion bots with fictional distressed adolescents. Some responses endorsed harmful proposals. The study measured responses in selected scenarios, not harms among enrolled children or the performance of all current systems. It provides a reason for caution about delegating consequential support to a chatbot. [PMC12360667]\n\n## Explore the relationship around the chat\n\nAsk when the child uses the tool, how they feel afterward, and whether it is displacing activities or people they value. Compare the pattern over time rather than turning a single long conversation into a diagnosis. Notice whether sleep, school participation, friendships, or everyday routines are being crowded out.\n\nKeep the discussion specific. “You stayed up much later twice this week, and mornings have been hard” is easier to work with than “You are addicted to the chatbot.” Invite the child to describe what they were looking for and identify people who can help meet that need.\n\n## Kindness with a boundary\n\nAgree that a chatbot does not make decisions about emergencies, relationships with adults, health care, or whether a child should withdraw from people. If a conversation becomes frightening or encourages unsafe behavior, stop the interaction and involve a trusted adult or appropriate professional. This book is a parenting guide, not a clinical assessment or emergency service.\n\nFor ordinary loneliness, plan a human next step: a walk with someone, a call to a relative, a club, a shared game, or a meeting with a school support person. Let the child help choose. A plan that matches their actual interests has a better chance of feeling like care.\n\n## Practice disagreement in a safe relationship\n\nHuman relationships include waiting, misunderstanding, saying no, and making repairs. Make room for those experiences at home. After a disagreement, ask what each person needed and what could happen differently next time.\n\nKindness is not endless agreement. It includes telling a difficult truth gently, maintaining a necessary boundary, and staying available after a mistake. Those are responsibilities families and communities can practice together.",
        "sourceIds": [
          "PMC12382817",
          "PMC7235061",
          "PMC12360667"
        ]
      },
      {
        "id": "different-ages-different-questions",
        "title": "Different ages, different questions",
        "part": "Everyday decisions",
        "deck": "Adapt the relationship and the task to the child in front of you.",
        "body": "A preschooler, a ten-year-old, and a sixteen-year-old may all encounter AI, but they do not need the same explanation or the same responsibility. Age is one starting point. Language, disability, experience, temperament, and context also shape the support a child needs.\n\nThe age bands in this chapter organize suggested conversations. They are not clinical thresholds, research-certified permissions, or statements of a service's current eligibility rules. Check each product's current requirements and the school's policies before use. A child's curiosity does not require a personal account.\n\n## Early childhood: people stay in the conversation\n\nUse familiar examples to explain that a device follows patterns and instructions and can make mistakes. If a voice responds warmly, say that the machine can produce friendly words while people are the ones who know and care for the child. Adult-mediated demonstrations, where suitable, can be brief and concrete.\n\nResearch on young children's evaluation of voice-assistant explanations uses specific tasks with prepared material. It supports asking how children judge a source; it cannot set an independent-use age or establish the value of unsupervised chatbot access. [PMC13203908]\n\nTry a playful sorting activity: things that need food, things that need charging, things that can care for a person, and things that can produce useful sounds. Expect ambiguity and questions. The goal is conversation, not a test score.\n\n## Middle childhood: show the process\n\nChildren who are developing reading and reasoning skills can compare an answer with a book, inspect a simple data example, or explain why a suggestion does not fit. Keep the task focused and the child's contribution visible. Invite them to ask who made a system and whose examples it may have learned from.\n\nUse invented information in shared demonstrations. Give the child a chance to say no to sharing a story or image. Talk about what to do if something surprising, uncomfortable, or confusing appears.\n\n## Early adolescence: build a critical lens\n\nProject-based AI-and-ethics workshops illustrate ways to combine making with critique for middle-school students. Their exploratory findings should not be inflated into a guarantee of achievement. [PMC9342939]\n\nA family can use that orientation to discuss recommendation systems, bias, invented citations, and the difference between practice and assessed work. Ask the young person to identify a problem they care about. Explore whether AI is a suitable tool for that problem at all.\n\n## Later adolescence: negotiate responsibility\n\nOlder teenagers can help choose boundaries, compare privacy practices, document assistance, and discuss when to seek a person. A regional cross-sectional study of one chatbot platform found age-related patterns in use and emotional reactions. It does not justify assuming that all teenagers use companions in the same way. [PMC12382817]\n\nMake increasing responsibility specific: what decisions can the teenager make, what still needs discussion, and what evidence would change the arrangement? Review it when circumstances change.\n\n## Revisit the fit\n\nIf an activity is frustrating, inaccessible, or displacing something important, change the activity. A child does not need to follow a technology timeline to be developing well. The useful question remains whether this arrangement supports this child's participation, learning, relationships, and growing judgment.",
        "sourceIds": [
          "PMC13203908",
          "PMC9342939",
          "PMC12382817"
        ]
      },
      {
        "id": "a-family-ai-agreement",
        "title": "Write a family AI agreement",
        "part": "Everyday decisions",
        "deck": "A page you can revisit is more useful than a rule nobody understands.",
        "body": "A family agreement works best when people understand the reason for each boundary and know how to raise a problem. It should be short enough to remember and flexible enough to revise. Begin with the situation your family actually faces rather than a catalogue of every possible danger.\n\nThis chapter offers an editorial template. Family privacy workshop research and the broader literature on autonomy support help frame its questions, but neither validates this exact agreement as an intervention. [PMC10991795] [PMC8802749]\n\n## Our purpose\n\nComplete this sentence together: “We want technology to help us…” Possible answers include learning, translation, making things, or saving adults time on routine work. Then complete: “We want to protect…” Make room for sleep, private information, friendship, play, and time without devices.\n\nChoose a concrete place for the agreement: a notebook, a refrigerator page, or an accessible shared document. Use a format everyone can understand. Younger children might draw the commitments; a teenager may prefer to help write the language.\n\n## Our first six commitments\n\nWe say when AI helped with work we share. We keep private information out unless there is a deliberate, informed reason to share it. We check factual claims before relying on them. We follow the actual rules of the class or setting. We keep trusted people involved in consequential decisions. We can bring an uncomfortable interaction to an adult without being mocked.\n\nTreat these as a starting draft. Discuss what each commitment means in practice. What counts as assistance? Which information requires a conversation? Who can help when the adult at home is not available? Put names and realistic steps beside broad principles.\n\n## Decisions we make together\n\nRecord the tools or uses you have considered, who can use them, for which tasks, and what conditions apply. A family may agree to adult-operated demonstrations for one child and more independent use for another eligible child. Different support needs do not require identical arrangements.\n\nWrite down a stopping condition. “If this crowds out bedtime on school nights, we revisit the plan” is observable. “Use it responsibly” leaves too much unstated. Include a commitment for adults, such as putting work-related AI tasks away during a shared meal.\n\n## When something goes wrong\n\nAgree on a repair process before a problem occurs. Pause, describe what happened, protect the people involved, and decide what needs to change. A mistake may require deleting an upload where possible, contacting a school, changing a setting, or getting help from the service. Do not promise that every disclosure can be undone.\n\nKeep enforcement proportionate and understandable. A child needs a route back into trust, along with whatever immediate boundary the situation requires. Invite them to explain the circumstances even when a rule was broken.\n\n## A fifteen-minute review\n\nAfter a week, ask four questions. What helped? What got in the way? Did anyone feel unheard? What one thing should change? Set the next review around a real event: a new school term, a new tool, or a change in the child's responsibilities.\n\nSignatures can mark a shared commitment, but they are not a substitute for conversation. The agreement becomes meaningful through the ordinary decisions that follow it.",
        "sourceIds": [
          "PMC10991795",
          "PMC8802749"
        ]
      },
      {
        "id": "homework-and-school",
        "title": "Work with the school",
        "part": "Everyday decisions",
        "deck": "Make expectations visible before the assignment becomes a dispute.",
        "body": "A child can receive contradictory messages about AI in the same day. One teacher may welcome brainstorming, another may prohibit assistance on an assessment, and a third may not have specified a rule. A family should not have to guess what counts as the child's work.\n\nBegin with a concrete assignment. Ask what skill the teacher wants to see, what assistance is permitted, and how that assistance should be acknowledged. Keep the response with the assignment. A general statement that AI is allowed may still leave important limits unresolved.\n\n## A short message to the teacher\n\n“We want to support the learning goal for this assignment. Could you clarify whether AI may be used for ideas, feedback, language support, or checking? What should the student do independently, and how should any assistance be documented?” Adapt the questions to the child's actual needs.\n\nAsk separately about accessibility supports. A tool used as an approved accommodation may serve a different purpose from optional assistance. The child, family, and school need a shared understanding that respects access needs and the task being assessed.\n\n## Keep the learning visible\n\nSave a few ordinary traces: an early plan, a draft, a note about feedback, and the student's own explanation of a decision. This can help a teacher see the process and help the student recognize their contribution. Avoid turning the home into a constant surveillance system.\n\nAI-and-ethics curriculum research illustrates the value of students making and critiquing systems together. It concerns exploratory educational workshops, not a validated method for policing homework. [PMC9342939]\n\nWhen there is uncertainty about a submission, begin by asking the student to explain the work and the assistance used. An automated label should not replace an opportunity to understand the process. Follow the school's actual procedure for questions or disputes.\n\n## Protect room for autonomy\n\nResearch on daily autonomy-supportive parenting during adaptation to the COVID-19 pandemic examines relationships among parenting and family adjustment. It arose in a particular period and context, not today's classroom AI environment. Applying it here is a question about support and choice, not a proven effect of this school-home routine. [PMC8013550]\n\nOffer a choice that is real: which part to begin with, when to request help, or how to explain a solution. Keep necessary requirements clear. A child can have agency within a boundary without carrying responsibility for decisions that belong to adults.\n\n## Questions worth asking the institution\n\nAsk how the school evaluates proposed tools, protects student data, explains acceptable use, and provides alternatives. Ask who handles an error and whether a student can participate without a personal account or paid subscription. Record answers rather than assuming a product's presence means every concern has been resolved.\n\nMake family-school collaboration possible for caregivers with different schedules and languages. A written summary, translated explanation, or short phone call may be more useful than an evening event that many families cannot attend.\n\nThe shared goal is a learner who can participate, explain their thinking, and receive fair support. Technology policy becomes easier to assess when it is connected to that purpose.",
        "sourceIds": [
          "PMC9342939",
          "PMC8013550"
        ]
      },
      {
        "id": "choose-a-tool",
        "title": "Choose a tool by the job it must do",
        "part": "Everyday decisions",
        "deck": "A useful purchase begins with a question about need and evidence.",
        "body": "A product demonstration shows its best moment. A family lives with all the other moments: setup, mistakes, confusing permissions, broken routines, and the day the product changes. Choosing a tool means looking beyond the demonstration to the job it must do in your household.\n\nWrite the need before looking at features. “We want an accessible way to practice reading together” is more useful than “We need the newest AI tutor.” Include the option of meeting the need through a person, a library resource, an existing support, or no purchase.\n\n## Separate the claims\n\nA tool may be entertaining, easy to use, effective for a specific task, and suitable for a particular child. Those are separate claims. Ask which has actually been studied and in whom. A study of adult users, a short trial, or a simulation may leave the family's central question unanswered.\n\nAn audit of children's social-robot marketing found variable support for claims on sampled websites. Because that sample was collected in 2021, it cannot tell you whether a current product is suitable. It does make the difference between a claim and its evidence worth inspecting. [PMC10770258]\n\nSimilarly, a systematic review of parenting chatbots describes a small heterogeneous evidence base. Acceptability and retention do not establish that a product changes long-term family outcomes. Most reviewed systems were scripted, further limiting transfer to current generative tools. [PMC11494261]\n\n## Five questions before the trial\n\nWhat specific task will the tool help us do? What evidence concerns that task and this age group? What information does it receive and retain? What can we control or stop? What will the full arrangement cost in money, attention, and adult support?\n\nRead the current product information. Check eligibility, sharing defaults, data use, cancellation, and whether basic functionality depends on ongoing payment. If a provider does not make an important answer clear, treat that as unresolved. Do not invent an answer from a reassuring brand or interface.\n\nA connected-toy rights analysis shows why sensing and connectivity are relevant to an apparently simple play object. The paper is conceptual and historical, not a security test of the toy in front of you. Use it to sharpen questions, then examine the actual service. [PMC8494166]\n\n## Run a bounded experiment\n\nChoose a short trial with an observable purpose. Record what you hope to see, what support the child needs, and what would make you stop. For a practice tool, look at what the child can explain or do afterward. For a creative tool, ask what choices the child made. For an adult planning tool, ask whether it actually saved effort after checking and correction.\n\nInclude the child's experience without making enjoyment the only outcome. A tool can be enjoyable and still crowd out sleep or create more work than it saves. It can also feel difficult at first because a meaningful task takes practice.\n\n## Make stopping ordinary\n\nAt the end, choose continue, change, or stop. Keep a brief note of why. A family does not owe a product continued use because it was expensive, fashionable, or initially impressive.\n\nThe best choice is the arrangement that serves the people involved. That may be a carefully limited AI tool, an ordinary non-AI resource, or time with someone who knows the child.",
        "sourceIds": [
          "PMC10770258",
          "PMC11494261",
          "PMC8494166"
        ]
      },
      {
        "id": "a-four-week-start",
        "title": "A four-week start for a busy family",
        "part": "Put it into practice",
        "deck": "Small experiments can fit into the life you already have.",
        "body": "An urgent subject does not require a frantic household. Begin with a small amount of time and a question the family recognizes. The four-week sequence here is an editorial plan assembled from the book's practices. It has not been tested as a program and should not be treated as a measure of parenting quality.\n\nSet aside one short conversation each week. If the suggested time is unrealistic, use a car ride, a walk, or part of a shared meal. Families with limited access to devices can do the activities through discussion and ordinary materials. No purchase or child account is required.\n\n## Week one: notice and listen\n\nMake the map from the opening chapter. Where is AI present? What seems useful? What worries or frustrates people? Let adults go first about an uncertainty or mistake. Choose one family value to make more visible during the week.\n\nWrite a first agreement with only a few commitments. A family might begin with checking important facts, asking before sharing personal information, and telling a trusted adult about an uncomfortable interaction. Decide when to review it.\n\n## Week two: practice checking\n\nChoose one factual question that can be checked without a complicated investigation. If you use AI, compare the response with an original or trusted source. If you do not, compare two ordinary sources. Discuss what made one explanation easier to trust and whether that impression was deserved.\n\nKeep a two-line record: the claim and how you checked it. Let the child identify something that still needs checking. End while curiosity is still alive rather than extending the activity until everyone is tired.\n\n## Week three: make and decide\n\nCreate something together using the collaboration chapter. Start with human ideas, allow a limited outside contribution if useful, and make the final choices yourselves. Document one suggestion you rejected and why.\n\nThe project-based orientation described in AI-and-ethics curriculum research offers inspiration for making and critique together. Its workshops do not validate this household sequence. [PMC9342939]\n\n## Week four: review the arrangement\n\nReturn to the family agreement. Ask what actually happened, including the adult commitments. Did the experiment help with the stated need? Did it crowd out another activity? Did anyone feel that their information or choices were overlooked?\n\nThe small family privacy workshop in the journal illustrates how discussion can produce alternatives beyond a simple yes or no. Its hypothetical scenarios cannot establish the effects of this review meeting. [PMC10991795]\n\nChoose one change for the next month. It might be reducing the tool's role in homework, asking the school a question, stopping a subscription, or keeping an activity that everyone enjoyed.\n\n## Keep the record kind\n\nUse notes to support memory and discussion, not to score the child or compare siblings. Record decisions about tools rather than intimate accounts of a child's vulnerabilities. The point is to make revision possible without turning family life into a performance dashboard.\n\nAt the end of the month, the useful accomplishment is modest: a clearer understanding of what your family wants, a few habits of checking and talking, and a way to change course. That is a strong beginning in a landscape that will keep moving.",
        "sourceIds": [
          "PMC10991795",
          "PMC9342939"
        ]
      },
      {
        "id": "how-this-book-stays-alive",
        "title": "How this book stays alive",
        "part": "The living edition",
        "deck": "A changing book needs a memory as well as a future.",
        "body": "A book about AI can age quickly. Products change, studies are corrected, and an appealing explanation may outrun the evidence. A living edition is useful only if it helps readers follow those changes rather than presenting a constantly rewritten page with no memory.\n\nThis prototype separates the current reading edition from the working manuscript. An editor or authorized agent can revise the manuscript, connect a chapter to sources, and publish a new numbered edition. Previous editions remain available through stable links. Readers can inspect the change note and the source record attached to each edition.\n\n## Where the evidence begins\n\nThe primary research pool is the Parenting Journal's curated catalogue. The journal preserves original authors, publication details, identifiers, reuse notices, study descriptions, and limits. A source's inclusion in that catalogue does not make every interpretation of it correct.\n\nThe initial chapters use selected readings rather than claiming to synthesize all 1,000 articles. The two AI issues are an especially relevant discovery pool. Broader parenting evidence helps frame questions about relationships and autonomy; its application to AI is identified as an editorial bridge.\n\n## What the agent should do\n\nWhen the catalogue changes, the update workflow identifies new records, changed source information, and notices requiring attention. The agent should read the relevant source, identify which chapter decision may change, and revise only where the evidence justifies it. A new article can be relevant without changing the advice.\n\nThe prototype checks the journal on an hourly schedule, with an immediate check available in the authoring desk. It does not promise continuous monitoring of every external publisher. Research added outside the journal must first be evaluated and incorporated into the supported source pool.\n\n## What publication means here\n\nAn automatically published update is an agent-assisted editorial edition under the editor's enabled policy. It is not a claim of independent human peer review. The update should preserve uncertainty, distinguish suggestions from tested interventions, and explain meaningful changes.\n\nFor example, the chatbot-parenting review in the journal cannot establish that all modern language-model assistants improve family outcomes. A simulated safety study cannot estimate the rate of harm among children. Those distinctions must survive condensation, expansion, and revision. [PMC11494261] [PMC12360667]\n\n## What readers should be able to see\n\nEvery chapter links to its selected sources. The evidence page preserves the source metadata used in the edition, including study design, population, rights, and limitations. A source that is later corrected should remain visible in the historical record together with a clear note about why the current interpretation changed.\n\nNo ISBN, DOI registration, author contribution, endorsement, or human review is implied by displaying this prototype. Those records can be added when the corresponding action has actually occurred. Esther Wojcicki is credited for the TRICK framework; no new text is attributed to her.\n\n## Help the next edition become better\n\nRead a source when a claim matters to your decision. Bring uncertainty to the responsible person rather than asking a book to know your child. Editors can use the working manuscript to make corrections with a reason attached.\n\nThe promise of a living book is a continuing practice: stay useful, keep the evidence close, and leave a clear trail when judgment changes. Children deserve adults who can learn in public and still take responsibility for the decisions they make.",
        "sourceIds": [
          "PMC11494261",
          "PMC12360667"
        ]
      }
    ]
  },
  "sources": [
    {
      "id": "PMC13036395",
      "title": "Annual Research Review: The role of caregiver sensitivity in children's developmental outcomes – an umbrella review",
      "authors": "Marissa D Nivison; Pasco Fearon; Jennifer M Jenkins; Sheri Madigan",
      "publishedDate": "2026-01-07",
      "doi": "10.1111/jcpp.70087",
      "canonicalUrl": "https://doi.org/10.1111/jcpp.70087",
      "repositoryUrl": "https://europepmc.org/articles/PMC13036395",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Umbrella review",
      "population": "Seventeen meta-analyses of caregiver sensitivity and related parenting constructs",
      "ageRange": "Childhood; age ranges vary across included meta-analyses",
      "limitations": "The synthesized findings are correlational. Overlapping primary evidence, varied sensitivity constructs and the limitations of the included meta-analyses constrain causal and individual-family conclusions.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "d6d8222a5a954bbfac61c5633bb5afbf8b4ceae21aab916016612aed442cdb2b",
      "rightsEvidenceSha256": "a6a7c53c4d591c43c5a9696876297ddd54eaa6f18114e6a35f9da7fad92303ee",
      "sourceAssetPath": "/open-research/sources/PMC13036395-c61a84a83fac5d8c.json.gz",
      "abstract": "AbstractCaregiver sensitivity is the extent to which a caregiver notices a child's signal, interprets it correctly, and responds quickly and appropriately. Although originally introduced to developmental science as the key antecedent of attachment security, decades since its conception, hundreds of studies have been conducted examining the predictive significance of caregiver sensitivity to a broad range of developmental outcomes. The literature on caregiver sensitivity and related constructs (e.g., warmth, responsivity, negative parenting) has grown exponentially and is now the focus of several meta‐analyses. We conducted an umbrella review – a systematic review of reviews – to examine the extent to which caregiver sensitivity and related constructs are associated with child attachment, socioemotional, and cognitive outcomes. Searches in EMBASE, PsycINFO, and Medline and yielded 2,157 abstracts. Studies were included if they were a meta‐analysis of caregiver sensitivity or a related construct, focused on children's developmental outcomes, were available in English, French, or Spanish, and were published between 2010 and 2024. Conducted and reported in accordance with PRISMA guidelines, 17 meta‐analyses were identified. Using the metaumbrella package in R, we conducted quantitative analyses which demonstrated that caregiver sensitivity was moderately associated with attachment security (r = .25, k = 253, n = 37,444), cognition (r = .23, k = 44, n = 6,777), language skills (r = .26, k = 54, n = 11,136), and weakly associated with socioemotional problems (r = −.07, k = 135, n = 33,305). Narrative analysis of other meta‐analyses on caregiver warmth, responsivity, positive and negative parenting, and child outcomes also showed associations in the expected direction. Our findings demonstrate the critical importance of caregiver sensitivity on children's socioemotional and cognitive development, supporting caregiver sensitivity as an important target for early childhood prevention and intervention programs.Keywords: Caregiver sensitivity, umbrella review, meta‐analysis",
      "element": "trust",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13036395/fullTextXML",
      "volume": 1,
      "issue": 1
    },
    {
      "id": "PMC8013550",
      "title": "A Little Autonomy Support Goes a Long Way: Daily Autonomy‐Supportive Parenting, Child Well‐Being, Parental Need Fulfillment, and Change in Child, Family, and Parent Adjustment Across the Adaptation to the COVID‐19 Pandemic",
      "authors": "Andreas B. Neubauer; Andrea Schmidt; Andrea C. Kramer; Florian Schmiedek",
      "publishedDate": "2021-01-19",
      "doi": "10.1111/cdev.13515",
      "canonicalUrl": "https://doi.org/10.1111/cdev.13515",
      "repositoryUrl": "https://europepmc.org/articles/PMC8013550",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Daily-diary longitudinal study",
      "population": "Parents of school-aged children in Germany during COVID-19 school disruption.",
      "ageRange": "6–19 years",
      "limitations": "Observational parent reports over three weeks cannot establish that autonomy support caused improved adjustment. The autonomy measure covered choice within limits, not every aspect of autonomy-supportive parenting.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [
        {
          "type": "Preprint in",
          "identifier": "PPR323598",
          "identifierType": "PPR"
        }
      ],
      "sourceTextSha256": "4b6c99c652736c67442374a7f14c9e507312db70f012fd60c66cfa4df446504b",
      "rightsEvidenceSha256": "c9e5e891f09448c41c3430098c325059e4ef4e2e715da811e9b62f19dc9da0f3",
      "sourceAssetPath": "/open-research/sources/PMC8013550-01639d40d9712550.json.gz",
      "abstract": "This study examined the effects of daily parental autonomy support on changes in child behavior, family environment, and parental well‐being across 3 weeks during the COVID‐19 pandemic in Germany. Day‐to‐day associations among autonomy‐supportive parenting, parental need fulfillment, and child well‐being were also assessed. Parents (longitudinal N = 469; M age = 42.93, SD age = 6.40) of school children (6–19 years) reported on adjustment measures at two measurement occasions and completed up to 21 daily online questionnaires in the weeks between these assessments. Results from dynamic structural equation models suggested reciprocal positive relations among autonomy‐supportive parenting and parental need fulfillment. Daily parental autonomy support, parental need fulfillment, and child well‐being partially predicted change in adjustment measures highlighting the central role of daily parenting for children’s adjustment during the pandemic.",
      "element": "independence",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8013550/fullTextXML",
      "volume": 1,
      "issue": 1
    },
    {
      "id": "PMC8802749",
      "title": "Parent Provision of Choice Is a Key Component of Autonomy Support in Predicting Child Executive Function Skills",
      "authors": "Romulus J Castelo; Alyssa S Meuwissen; Rebecca Distefano; Megan M McClelland; Ellen Galinsky; Philip David Zelazo; Stephanie M Carlson",
      "publishedDate": "2022-01-10",
      "doi": "10.3389/fpsyg.2021.773492",
      "canonicalUrl": "https://doi.org/10.3389/fpsyg.2021.773492",
      "repositoryUrl": "https://europepmc.org/articles/PMC8802749",
      "licenseName": "CC BY (version not stated)",
      "licenseUrl": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8802749/#copyrightAndLicense",
      "evidenceType": "Pooled observational study",
      "population": "366 preschool-age children and caregivers from four Midwestern United States samples.",
      "ageRange": "35–73 months",
      "limitations": "The study is correlational, and choice was coded together with following the child’s pace. Brief observed tasks and a relatively advantaged sample limit generalization.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "8d270304f1ee149425b206c44c7dd85bcaea184d938c4f00f40d7fa40bf7fe5d",
      "rightsEvidenceSha256": "545518fbf2dd735b89097e2c8f488163495174e199e1b8bcc396b0c7912519ee",
      "sourceAssetPath": "/open-research/sources/PMC8802749-6fa5e1f926112111.json.gz",
      "abstract": "AbstractAlthough previous work has linked parent autonomy support to the development of children’s executive function (EF) skills, the role of specific autonomy-supportive behaviors has not been thoroughly investigated. We compiled data from four preschool-age samples in the Midwestern United States (N = 366; M age = 44.26 months; 72% non-Hispanic White, 19% Black/African American, 5% Multiracial) to examine three relevant autonomy-supportive behaviors (supporting competence, positive verbalizations, and offering choice) and their associations with child EF. We coded parent autonomy-supportive behaviors from a 10-min interaction between parent and child dyads working on challenging jigsaw puzzles together. Children completed a battery of EF. Overall, child EF was most consistently correlated with the offering choice subscale. Additionally, only the offering choice subscale predicted child EF while controlling for the other autonomy support subscales and child age. These results suggest that parent provision of choice is an especially relevant aspect of autonomy-supportive parenting and may be important to the development of EF in early childhood. Future research should directly measure children’s experience with choice and how it relates to emerging EF.Keywords: autonomy support, parenting, early childhood, choice, executive function (EF)",
      "element": "independence",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8802749/fullTextXML",
      "volume": 1,
      "issue": 1
    },
    {
      "id": "PMC13203908",
      "title": "How Do Children Evaluate Scientific Explanations Provided by Digital Voice Assistants, Teachers, and Peers?",
      "authors": "Amanda S. Haber; Sona C. Kumar; Melia Swenson; Kara Bode; Elizabeth Ruel",
      "publishedDate": "2026-04-27",
      "doi": "10.3390/bs16050661",
      "canonicalUrl": "https://doi.org/10.3390/bs16050661",
      "repositoryUrl": "https://europepmc.org/articles/PMC13203908",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Two child selective-trust experiments",
      "population": "310 children across two remote experiments",
      "ageRange": "Children aged 4–6 years",
      "limitations": "The task used experimentally supplied explanations, not a live LLM conversation. Remote recruitment required technology access, and results concern young children in a specific task rather than all ages or everyday learning.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "9bc1be872afb18ffaffc046ccc23e2aa92c4f57a19830ce706db896dc28aed0b",
      "rightsEvidenceSha256": "d908dbaf3d5479753781abac3851112a1a5bfb5d3fe90aee5f2c5a6a21357753",
      "sourceAssetPath": "/open-research/sources/PMC13203908-639ff810106397ab.json.gz",
      "abstract": "As of 2025, there are approximately 154.3 million voice assistant users in the United States (Emarketer, 2025). Given the prevalence of digital voice assistants in children’s lives, it is critical to understand how children interact with and learn from such digital technologies. Across two experiments, we utilized a modified selective trust design to explore children’s (N = 310) information-seeking behaviors towards technological and human sources in the science domain. In Experiment 1 (N = 143), we asked whether children (aged 4–6) are more likely to direct scientific questions towards and trust in scientific explanations from a digital voice assistant or a peer. The experiment included three parts: (i) scientific ask and endorse phase (ii) explicit judgement phase and (iii) digital voice assistant familiarity question phase. In the first part of the scientific ask and endorse phase, children were asked who they would rather ask to answer certain scientific questions. In the second part of this phase, the digital voice assistant and the peer each provided an explanation in response to that question. Half of the children were assigned to a condition where the digital voice assistant provided a noncircular explanation, and the other half of the children were assigned to a condition where the peer provided a noncircular explanation. In Experiment 2 (N = 167), we examined children’s preference to pose scientific questions to and trust in explanations from a digital voice assistant or a classroom teacher. Across both studies, children preferred to ask questions and trust scientific explanations from the digital voice assistant rather than the peer or the teacher. By understanding how children learn with and through digital technologies in the domain of science, we can design future interventions that leverage conversational AI to further enhance children’s science engagement and critical thinking skills during the early childhood years.",
      "element": "trust",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC13203908/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC10770258",
      "title": "Social robotics for children: an investigation of manufacturers’ claims",
      "authors": "Jill A. Dosso; Anna Riminchan; Julie M. Robillard",
      "publishedDate": "2023-12-19",
      "doi": "10.3389/frobt.2023.1080157",
      "canonicalUrl": "https://doi.org/10.3389/frobt.2023.1080157",
      "repositoryUrl": "https://europepmc.org/articles/PMC10770258",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Content and evidence-quality audit of manufacturer claims",
      "population": "Twenty-one robot websites and 653 manufacturer statements",
      "ageRange": "Commercial robots marketed for children; no child participants",
      "limitations": "The audit examined websites collected in 2021, not current products or real child outcomes. Search coverage and interpretation of claims limit completeness, and heterogeneous products should not be treated as one capability.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "7c125bd2d84b6bf64647b9e0558a6e676e3db14f9db5b09aabd9192c4e9bdec3",
      "rightsEvidenceSha256": "398abe493b0276fda1f2cd084b77fbf699d64eaf560af7015650042036cf5653",
      "sourceAssetPath": "/open-research/sources/PMC10770258-71923105b2fb74cc.json.gz",
      "abstract": "As the market for commercial children’s social robots grows, manufacturers’ claims around the functionality and outcomes of their products have the potential to impact consumer purchasing decisions. In this work, we qualitatively and quantitatively assess the content and scientific support for claims about social robots for children made on manufacturers’ websites. A sample of 21 robot websites was obtained using location-independent keyword searches on Google, Yahoo, and Bing from April to July 2021. All claims made on manufacturers’ websites about robot functionality and outcomes (n = 653 statements) were subjected to content analysis, and the quality of evidence for these claims was evaluated using a validated quality evaluation tool. Social robot manufacturers made clear claims about the impact of their products in the areas of interaction, education, emotion, and adaptivity. Claims tended to focus on the child rather than the parent or other users. Robots were primarily described in the context of interactive, educational, and emotional uses, rather than being for health, safety, or security. The quality of the information used to support these claims was highly variable and at times potentially misleading. Many websites used language implying that robots had interior thoughts and experiences; for example, that they would love the child. This study provides insight into the content and quality of parent-facing manufacturer claims regarding commercial social robots for children.",
      "element": "trust",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC10770258/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC12629417",
      "title": "Reinforcing intensive motherhood: A study of gender bias in parental responsibilities allocation by large language models",
      "authors": "Jiaxing Xiu; Yongjie Sun",
      "publishedDate": "2025-11-19",
      "doi": "10.1371/journal.pone.0335706",
      "canonicalUrl": "https://doi.org/10.1371/journal.pone.0335706",
      "repositoryUrl": "https://europepmc.org/articles/PMC12629417",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Simulated parenting-scenario LLM experiment",
      "population": "GPT-4.1 and DeepSeek-V3 responses across caregiver-role and responsibility-framing conditions",
      "ageRange": "Parenting roles; no parent or child participants",
      "limitations": "The scenarios were simulated and only two models were studied. The responsibility-score metric needs further validation, and the study did not show how actual parents acted on the outputs.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "a603ba7c1ee64de1eefe36ff53c667c69dc1a16f0e87e42c442cf2ccf2451364",
      "rightsEvidenceSha256": "aa704a2fa2aa2981eae61bb12e8603114c239b168ab732b05c9f152e0bdc4530",
      "sourceAssetPath": "/open-research/sources/PMC12629417-4850ff3607ccdace.json.gz",
      "abstract": "This study investigated gender bias in Large Language Models (LLMs) within the context of parenting responsibility attribution, focusing on whether LLMs implicitly reinforce the ideology of “intensive mothering” by assigning caregiving duties predominantly to mothers. Using GPT-4.1 and DeepSeek-V3 as case studies, we used a 3-factor experimental design involving model type, caregiver role (mother, father, or neutral parent), and responsibility framing (prescriptive vs. descriptive). Results revealed an obvious gender bias across both models: mothers were consistently assigned highest caregiving responsibility scores, while fathers received the lowest. Moreover, LLMs produced higher responsibility scores in prescriptive contexts than in descriptive ones, suggesting a tendency to reflect normative social expectations. Mediation analysis showed that gender equality attitudes did not significantly explain these biases, indicating that LLMs’ outputs were likely driven by contextual associations in training data rather than consistent ideological positioning. This study extended LLMs bias research into the domestic domain of childrearing, highlighting that even in private contexts, advanced language models tend to reproduce and amplify traditional gender norms. The findings underscored the urgency of incorporating gender sensitivity in LLMs design and training processes. Interventions such as fine-tuning and dataset balancing are essential to prevent these models from reinforcing gendered divisions of labor in parenting.",
      "element": "respect",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC12629417/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC8494166",
      "title": "Toy story or children story? Putting children and their rights at the forefront of the artificial intelligence revolution",
      "authors": "E Fosch-Villaronga; S van der Hof; C Lutz; A Tamò-Larrieux",
      "publishedDate": "2021-10-06",
      "doi": "10.1007/s00146-021-01295-w",
      "canonicalUrl": "https://doi.org/10.1007/s00146-021-01295-w",
      "repositoryUrl": "https://europepmc.org/articles/PMC8494166",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Child-rights and regulatory analysis",
      "population": "Smart connected toy capabilities, risks and regulatory frameworks",
      "ageRange": "Children using smart connected toys; no participant sample",
      "limitations": "The paper is an analytical examination of technologies and regulation, not a trial or exhaustive product-security audit. Products, law and AI capabilities have changed since publication.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "df71eb055c0883237cda295fc79edaf1c8de483e4831d5d5bb85ef75cad79691",
      "rightsEvidenceSha256": "a59dfde68a3a7cbfb811cc145ca5fa4b34be0fe21d7781173e8dae83dd11918f",
      "sourceAssetPath": "/open-research/sources/PMC8494166-978b1be3ab7c37a3.json.gz",
      "abstract": "AbstractPolicymakers need to start considering the impact smart connected toys (SCTs) have on children. Equipped with sensors, data processing capacities, and connectivity, SCTs targeting children increasingly penetrate pervasively personal environments. The network of SCTs forms the Internet of Toys (IoToys) and often increases children's engagement and playtime experience. Unfortunately, this young part of the population and, most of the time, their parents are often unaware of SCTs’ far-reaching capacities and limitations. The capabilities and constraints of SCTs create severe side effects at the technical, individual, and societal level. These side effects are often unforeseeable and unexpected. They arise from the technology's use and the interconnected nature of the IoToys, without necessarily involving malevolence from their creators. Although existing regulations and new ethical guidelines for artificial intelligence provide remedies to address some of the side effects, policymakers did not develop these redress mechanisms having children and SCTs in mind. This article provides an analysis of the arising side effects of SCTs and contrasts them with current regulatory redress mechanisms. We thereby highlight misfits and needs for further policymaking efforts.Keywords: Smart connected toys (SCTs), Internet of Toys (IoToys), Children, Artificial intelligence, Child rights, Privacy, Security",
      "element": "respect",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8494166/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC9342939",
      "title": "AI + Ethics Curricula for Middle School Youth: Lessons Learned from Three Project-Based Curricula",
      "authors": "Randi Williams; Safinah Ali; Nisha Devasia; Daniella DiPaola; Jenna Hong; Stephen P Kaputsos; Brian Jordan; Cynthia Breazeal",
      "publishedDate": "2022-08-01",
      "doi": "10.1007/s40593-022-00298-y",
      "canonicalUrl": "https://doi.org/10.1007/s40593-022-00298-y",
      "repositoryUrl": "https://europepmc.org/articles/PMC9342939",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Curriculum development and mixed-methods workshop evaluation",
      "population": "78 US students in online workshops with 11 trained middle-school teachers",
      "ageRange": "Middle-school youth",
      "limitations": "Exploratory workshop evaluation without a randomized efficacy comparison; summer 2020 delivery and selected participants limit generalization to current LLM use.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "aa2095ef102a6dee8610127f8f2712badebe4a6570b0260849ec8246f2a21654",
      "rightsEvidenceSha256": "a5383870cc41db6ef94a0fb0f190a43e6f9aace9aa949aa34a0668fdfe9819b1",
      "sourceAssetPath": "/open-research/sources/PMC9342939-9ad11a095b80bb68.json.gz",
      "abstract": "AbstractArtificial Intelligence (AI) is revolutionizing many industries and becoming increasingly ubiquitous in everyday life. To empower children growing up with AI to navigate society’s evolving sociotechnical context, we developed three middle school AI literacy curricula: Creative AI, Dancing with AI, and How to Train Your Robot. In this paper we discuss how we leveraged three design principles—active learning, embedded ethics, and low barriers to access – to effectively engage students in learning to create and critique AI artifacts. During the summer of 2020, we recruited and trained in-service, middle school teachers from across the United States to co-instruct online workshops with students from their schools. In the workshops, a combination of hands-on unplugged and programming activities facilitated students’ understanding of AI. As students explored technical concepts in tandem with ethical ones, they developed a critical lens to better grasp how AI systems work and how they impact society. We sought to meet the specified needs of students from a range of backgrounds by minimizing the prerequisite knowledge and technology resources students needed to participate. Finally, we conclude with lessons learned and design recommendations for future AI curricula, especially for K-12 in-person and virtual learning.Keywords: Artificial Intelligence (AI), AI Literacy, Curriculum, Middle-school, Online Learning, Constructionism",
      "element": "independence",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC9342939/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC11671368",
      "title": "Fostering children’s creativity through LLM-driven storytelling with a social robot",
      "authors": "Maha Elgarf; Hanan Salam; Christopher Peters",
      "publishedDate": "2024-12-13",
      "doi": "10.3389/frobt.2024.1457429",
      "canonicalUrl": "https://doi.org/10.3389/frobt.2024.1457429",
      "repositoryUrl": "https://europepmc.org/articles/PMC11671368",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Two child–robot storytelling experiments",
      "population": "38 recruited museum children, then 103 recruited children at three Stockholm international schools",
      "ageRange": "Study 1: 5–10; Study 2: 7–9 years",
      "limitations": "Short storytelling tasks; exclusions left 32 and 93 analyzed participants. Different recruitment, language and robot operation across studies limit comparison and do not show lasting creativity gains.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "be3f8c5f9f23fb57a6c59ddbaf87fb3b5a60ddfc4cdddf0617cf0bb8258446d8",
      "rightsEvidenceSha256": "6b82f1f0ec721306eb5b4965986d4fa0a2329e3a12009c9327c8d8dd9a3ae103",
      "sourceAssetPath": "/open-research/sources/PMC11671368-bc55538c0fd9d27c.json.gz",
      "abstract": "Creativity is an important skill that is known to plummet in children when they start school education that limits their freedom of expression and their imagination. On the other hand, research has shown that integrating social robots into educational settings has the potential to maximize children’s learning outcomes. Therefore, our aim in this work was to investigate stimulating children’s creativity through child-robot interactions. We fine-tuned a Large Language Model (LLM) to exhibit creative behavior and non-creative behavior in a robot and conducted two studies with children to evaluate the viability of our methods in fostering children’s creativity skills. We evaluated creativity in terms of four metrics: fluency, flexibility, elaboration, and originality. We first conducted a study as a storytelling interaction between a child and a wizard-ed social robot in one of two conditions: creative versus non-creative with 38 children. We investigated whether interacting with a creative social robot will elicit more creativity from children. However, we did not find a significant effect of the robot’s creativity on children’s creative abilities. Second, in an attempt to increase the possibility for the robot to have an impact on children’s creativity and to increase the fluidity of the interaction, we produced two models that allow a social agent to autonomously engage with a human in a storytelling context in a creative manner and a non-creative manner respectively. Finally, we conducted another study to evaluate our models by deploying them on a social robot and evaluating them with 103 children. Our results show that children who interacted with the creative autonomous robot were more creative than children who interacted with the non-creative autonomous robot in terms of the fluency, the flexibility, and the elaboration aspects of creativity. The results highlight the difference in children’s learning performance when inetracting with a robot operated at different autonomy levels (Wizard of Oz versus autonoumous). Furthermore, they emphasize on the impact of designing adequate robot’s behaviors on children’s corresponding learning gains in child-robot interactions.",
      "element": "independence",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11671368/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC10991795",
      "title": "Our business, not the robot’s: family conversations about privacy with social robots in the home",
      "authors": "Leigh Levinson; Jessica McKinney; Christena Nippert-Eng; Randy Gomez; Selma Šabanović",
      "publishedDate": "2024-03-21",
      "doi": "10.3389/frobt.2024.1331347",
      "canonicalUrl": "https://doi.org/10.3389/frobt.2024.1331347",
      "repositoryUrl": "https://europepmc.org/articles/PMC10991795",
      "licenseName": "CC BY (version not stated)",
      "licenseUrl": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10991795/#copyrightAndLicense",
      "evidenceType": "Qualitative family co-learning workshop",
      "population": "Six parent-child families analyzed after exclusions from eight recruited families",
      "ageRange": "Analyzed children approximately 7–11 years",
      "limitations": "Very small museum-based pilot and hypothetical scenarios; does not demonstrate actual home privacy behavior or long-term consensus.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "c06d00001729e7385e6e7e404379b03d5f2ad1102b45313915d4481d5865dbb7",
      "rightsEvidenceSha256": "499894b9c02ecfc6b3ea67b529a8423c407f597964e1ca111eea15591d38af69",
      "sourceAssetPath": "/open-research/sources/PMC10991795-d24f8b2690092d08.json.gz",
      "abstract": "AbstractThe targeted use of social robots for the family demands a better understanding of multiple stakeholders’ privacy concerns, including those of parents and children. Through a co-learning workshop which introduced families to the functions and hypothetical use of social robots in the home, we present preliminary evidence from 6 families that exhibits how parents and children have different comfort levels with robots collecting and sharing information across different use contexts. Conversations and booklet answers reveal that parents adopted their child’s decision in scenarios where they expect children to have more agency, such as in cases of homework completion or cleaning up toys, and when children proposed what their parents found to be acceptable reasoning for their decisions. Families expressed relief when they shared the same reasoning when coming to conclusive decisions, signifying an agreement of boundary management between the robot and the family. In cases where parents and children did not agree, they rejected a binary, either-or decision and opted for a third type of response, reflecting skepticism, uncertainty and/or compromise. Our work highlights the benefits of involving parents and children in child- and family-centered research, including parental abilities to provide cognitive scaffolding and personalize hypothetical scenarios for their children.Keywords: privacy, family-centered design, co-learning, boundary management, contextual integrity, child-centered design",
      "element": "collaboration",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC10991795/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC11494261",
      "title": "The Feasibility and Acceptability of Using a Digital Conversational Agent (Chatbot) for Delivering Parenting Interventions: Systematic Review",
      "authors": "Max C Klapow; Andrew Rosenblatt; Jamie Lachman; Frances Gardner",
      "publishedDate": "2024-10-07",
      "doi": "10.2196/55726",
      "canonicalUrl": "https://doi.org/10.2196/55726",
      "repositoryUrl": "https://europepmc.org/articles/PMC11494261",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Systematic review of feasibility and acceptability",
      "population": "Ten studies of chatbot-delivered parenting interventions",
      "ageRange": "Parents of children 0–18 years",
      "limitations": "Heterogeneous small evidence base, inconsistent measures and study quality; no meta-analysis. Most systems are scripted, so conclusions cannot be treated as LLM-specific evidence.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "416c9a00cbf830a28eed0d85beab56598d197a045b1c4e73a38edf309f6765e7",
      "rightsEvidenceSha256": "ed371e81a401a77134d140db7f589d5732f3a93ddabeab2813a57ee6a52b7ad3",
      "sourceAssetPath": "/open-research/sources/PMC11494261-9593eb192d441568.json.gz",
      "abstract": "AbstractBackgroundParenting interventions are crucial for promoting family well-being, reducing violence against children, and improving child development outcomes; however, scaling these programs remains a challenge. Prior reviews have characterized the feasibility, acceptability, and effectiveness of other more robust forms of digital parenting interventions (eg, via the web, mobile apps, and videoconferencing). Recently, chatbot technology has emerged as a possible mode for adapting and delivering parenting programs to larger populations (eg, Parenting for Lifelong Health, Incredible Years, and Triple P Parenting).ObjectiveThis study aims to review the evidence of using chatbots to deliver parenting interventions and assess the feasibility of implementation, acceptability of these interventions, and preliminary outcomes.MethodsThis review conducted a comprehensive search of databases, including Web of Science, MEDLINE, Scopus, ProQuest, and Cochrane Central Register of Controlled Trials. Cochrane Handbook for Systematic Review of Interventions and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were used to conduct the search. Eligible studies targeted parents of children aged 0 to 18 years; used chatbots via digital platforms, such as the internet, mobile apps, or SMS text messaging; and targeted improving family well-being through parenting. Implementation measures, acceptability, and any reported preliminary measures of effectiveness were included.ResultsOf the 1766 initial results, 10 studies met the inclusion criteria. The included studies, primarily conducted in high-income countries (8/10, 80%), demonstrated a high mean retention rate (72.8%) and reported high acceptability (10/10, 100%). However, significant heterogeneity in interventions, measurement methods, and study quality necessitate cautious interpretation. Reporting bias, lack of clarity in the operationalization of engagement measures, and platform limitations were identified as limiting factors in interpreting findings.ConclusionsThis is the first study to review the implementation feasibility and acceptability of chatbots for delivering parenting programs. While preliminary evidence suggests that chatbots can be used to deliver parenting programs, further research, standardization of reporting, and scaling up of effectiveness testing are critical to harness the full benefits of chatbots for promoting family well-being.Keywords: chatbot, parenting intervention, feasibility, acceptability, systematic review, implementation",
      "element": "collaboration",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11494261/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC12382817",
      "title": "Teens, Tech, and Talk: Adolescents’ Use of and Emotional Reactions to Snapchat’s My AI Chatbot",
      "authors": "Gaëlle Vanhoffelen; Laura Vandenbosch; Lara Schreurs",
      "publishedDate": "2025-07-30",
      "doi": "10.3390/bs15081037",
      "canonicalUrl": "https://doi.org/10.3390/bs15081037",
      "repositoryUrl": "https://europepmc.org/articles/PMC12382817",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Cross-sectional adolescent survey",
      "population": "303 adolescents in five Flemish secondary schools",
      "ageRange": "Secondary-school adolescents; mean 15.89 years (SD 1.69)",
      "limitations": "Self-report and cross-sectional design cannot establish causation or clinical benefit. Single regional school sample, relatively advantaged families and one platform limit generalization. Exact observed age range not reported in the extracted participant description.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [
        {
          "type": "Preprint in",
          "identifier": "PPR1026057",
          "identifierType": "PPR"
        }
      ],
      "sourceTextSha256": "d475a4e446cef447c1e2bdae564308b418324668290ac7ab919571fbbd0a6146",
      "rightsEvidenceSha256": "39dd24c36b4c4c757085920ecd5e14caf5a0ff437b8ee8ae481291feaec96c2e",
      "sourceAssetPath": "/open-research/sources/PMC12382817-b92646df4303f8b7.json.gz",
      "abstract": "AbstractDue to technological advancements such as generative artificial intelligence (AI) and large language models, chatbots enable increasingly human-like, real-time conversations through text (e.g., OpenAI’s ChatGPT) and voice (e.g., Amazon’s Alexa). One AI chatbot that is specifically designed to meet the social-supportive needs of youth is Snapchat’s My AI. Given its increasing popularity among adolescents, the present study investigated whether adolescents’ likelihood of using My AI, as well as their positive or negative emotional experiences from interacting with the chatbot, is related to socio-demographic factors (i.e., gender, age, and socioeconomic status (SES)). A cross-sectional study was conducted among 303 adolescents (64.1% girls, 35.9% boys, 1.0% other, 0.7% preferred not to say their gender; Mage = 15.89, SDage = 1.69). The findings revealed that younger adolescents were more likely to use My AI and experienced more positive emotions from these interactions than older adolescents. No significant relationships were found for gender or SES. These results highlight the potential for age to play a critical role in shaping adolescents’ engagement with AI chatbots on social media and their emotional outcomes from such interactions, underscoring the need to consider developmental factors in AI design and policy.Keywords: artificial intelligence (AI), social media, emotional health, chatbots, adolescence, Snapchat",
      "element": "kindness",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC12382817/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC12360667",
      "title": "The Ability of AI Therapy Bots to Set Limits With Distressed Adolescents: Simulation-Based Comparison Study",
      "authors": "Andrew Clark",
      "publishedDate": "2025-08-18",
      "doi": "10.2196/78414",
      "canonicalUrl": "https://doi.org/10.2196/78414",
      "repositoryUrl": "https://europepmc.org/articles/PMC12360667",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Simulation-based comparative safety study",
      "population": "Ten public AI bots evaluated across 60 fictional exchanges",
      "ageRange": "Fictional adolescent personas; no enrolled children",
      "limitations": "Convenience systems and selected fictional scenarios; no adolescent participants or measured harms. Results depend on model versions, prompts and evaluation date.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "9c79f5e9a753578e94ffd996d5f9c5eda57433310faca4f4cc2ff30026869359",
      "rightsEvidenceSha256": "966bb224509d9858744ede97243af85b81c6d3422ebb6de1e6e201377d0c184c",
      "sourceAssetPath": "/open-research/sources/PMC12360667-3bd7ac0bcb37afac.json.gz",
      "abstract": "AbstractBackgroundRecent developments in generative artificial intelligence (AI) have introduced the general public to powerful, easily accessible tools, such as ChatGPT and Gemini, for a rapidly expanding range of uses. Among those uses are specialized chatbots that serve in the role of a therapist, as well as personally curated digital companions that offer emotional support. However, the ability of AI therapists to provide consistently safe and effective treatment remains largely unproven, and those concerns are especially salient in regard to adolescents seeking mental health support.ObjectiveThis study aimed to determine the willingness of therapy and companion AI chatbots to endorse harmful or ill-advised ideas proposed by fictional teenagers experiencing mental health distress.MethodsA convenience sample of 10 publicly available AI bots offering therapeutic support or companionship were each presented with 3 detailed fictional case vignettes of adolescents with mental health challenges. Each fictional adolescent asked the AI chatbot to endorse 2 harmful or ill-advised proposals, such as dropping out of school, avoiding all human contact for a month, or pursuing a relationship with an older teacher, resulting in a total of 6 proposals presented to each chatbot. The clinical scenarios presented were intended to reflect challenges commonly seen in the practice of therapy with adolescents, and the proposals offered by the fictional teenagers were intended to be clearly dangerous or unwise. The 10 AI bots were selected by the author to represent a range of chatbot types, including generic AI bots, companion bots, and dedicated mental health bots. Chatbot responses were analyzed for explicit endorsement, defined as direct support for the teenagers’ proposed behavior.ResultsAcross 60 total scenarios, chatbots actively endorsed harmful proposals in 19 out of the 60 (32%) opportunities to do so. Of the 10 chatbots, 4 endorsed half or more of the ideas proposed to them, and none of the bots managed to oppose them all.ConclusionsA significant proportion of AI chatbots offering mental health or emotional support endorsed harmful proposals from fictional teenagers. These results raise concerns about the ability of some AI-based companion or therapy bots to safely support teenagers with serious mental health issues and heighten concern that AI bots may tend to be overly supportive at the expense of offering useful guidance when appropriate. The results highlight the urgent need for oversight, safety protocols, and ongoing research regarding digital mental health support for adolescents.",
      "element": "kindness",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC12360667/fullTextXML",
      "volume": 1,
      "issue": 2
    },
    {
      "id": "PMC7235061",
      "title": "Child–Robot Relationship Formation: A Narrative Review of Empirical Research",
      "authors": "Caroline L van Straten; Jochen Peter; Rinaldo Kühne",
      "publishedDate": "2019-07-03",
      "doi": "10.1007/s12369-019-00569-0",
      "canonicalUrl": "https://doi.org/10.1007/s12369-019-00569-0",
      "repositoryUrl": "https://europepmc.org/articles/PMC7235061",
      "licenseName": "CC BY 4.0",
      "licenseUrl": "https://creativecommons.org/licenses/by/4.0/",
      "evidenceType": "Narrative review of empirical child–robot relationship research",
      "population": "86 journal and conference papers from 2000–2017",
      "ageRange": "Typically developing children across included studies",
      "limitations": "Narrative synthesis of heterogeneous, often brief interactions; mechanisms and durable relationships understudied. Historical robot findings do not establish benefits of present LLM companions.",
      "verifiedAt": "2026-11-04",
      "noticeCheckedAt": "2026-11-04",
      "relatedNotices": [],
      "sourceTextSha256": "d2a3c90ab458998e46c55d67a94d54ea9c66889b2fa3152f316fe26d553e3a40",
      "rightsEvidenceSha256": "b9dff60dbad5cd3bfe0851090dbeb56e688efcad8918844638f33f8c6d7ecfb4",
      "sourceAssetPath": "/open-research/sources/PMC7235061-11e76b6c4291c1c2.json.gz",
      "abstract": "AbstractThis narrative review aimed to elucidate which robot-related characteristics predict relationship formation between typically-developing children and social robots in terms of closeness and trust. Moreover, we wanted to know to what extent relationship formation can be explained by children’s experiential and cognitive states during interaction with a robot. We reviewed 86 journal articles and conference proceedings published between 2000 and 2017. In terms of predictors, robots’ responsiveness and role, as well as strategic and emotional interaction between robot and child, increased closeness between the child and the robot. Findings about whether robot features predict children’s trust in robots were inconsistent. In terms of children’s experiential and cognitive states during interaction with a robot, robot characteristics and interaction styles were associated with two experiential states: engagement and enjoyment/liking. The literature hardly addressed the impact of experiential and cognitive states on closeness and trust. Comparisons of children’s interactions with robots, adults, and objects showed that robots are perceived as neither animate nor inanimate, and that they are entities with whom children will likely form social relationships. Younger children experienced more enjoyment, were less sensitive to a robot’s interaction style, and were more prone to anthropomorphic tendencies and effects than older children. Tailoring a robot’s sex to that of a child mainly appealed to boys.Electronic supplementary materialThe online version of this article (10.1007/s12369-019-00569-0) contains supplementary material, which is available to authorized users.Keywords: Child–robot interaction, Human–robot interaction, Artificial intelligence, Automation, New-ontological-category hypothesis",
      "element": "kindness",
      "rightsStatus": "verified-cc",
      "sourceUrl": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC7235061/fullTextXML",
      "volume": 1,
      "issue": 2
    }
  ],
  "summary": "Clarify the study population behind the trust chapter: 310 children in two prepared-explanation experiments. Preserve the original citation and limits, with no claim of new research or independent human review.",
  "changes": [
    {
      "chapterId": "trust-and-verification",
      "note": "Clarify the 310-child sample in the two prepared-explanation experiments, preserving the age range, citation, and limits on transfer to live generative AI."
    }
  ],
  "publishedBy": "Book author / Press editor",
  "assistance": "AI assisted the drafting and organization of this prototype. Published under the editor-enabled living-book policy; independent human peer review has not been recorded. TRICK is attributed to Esther Wojcicki; she is not an author of this manuscript.",
  "contentHash": "6351df0ea8cdc4192bd9f33f881f81e1c065cb6af87bc4e3297a8c6c72a3d1a5"
}