1. Introduction and problem

In the 0–6 span—kindergarten, preschool, CENDI, nursery, childcare—a package of four artefacts now claims to count as communication with families. The first is the notice app: a channel that replaces the printed circular and delivers menus, absences, lice and the week’s photos. The second is the “what did my child do today” chatbot: a conversational interface that answers, by template or generative model, the question an adult would ask at the door or on the phone. The third is the automatic digest of the day: a summary that condenses check-in, nap, meals, photos and, sometimes, a generated sentence about “today’s learning.” The fourth is the AI-enabled portal: a dashboard that drafts “friendly” messages, translates, schedules sends and promises “engagement.” All four are visible, auditable and cheap in coordination time. They allow a setting to exhibit, to supervisors, networks or families themselves, that it “already communicates with artificial intelligence.” The enunciative leap is large: from notifying, condensing or generating to asserting that pedagogical communication is taking place. That leap is not authorised by the evidence on family–school partnerships in early childhood education, nor by the evidence on the digital channels that mediated them in 2021–2026.

The thesis of this article is restrictive. A notice chatbot or an automatic digest of the day does not constitute pedagogical family–school communication. Pedagogical communication is situated dialogue among adults responsible for the child; not a generated feed. Developmentally appropriate practice states this with institutional clarity: educators seek and maintain regular, frequent, two-way communication, recognising that forms may differ for each family, and they include informal conversations at drop-off and pick-up, conferences, and reciprocal technology-mediated exchanges (National Association for the Education of Young Children [NAEYC], 2020). Starting Strong VI anchors the quality of early childhood education in process quality: everyday interactions with adults, peers, materials and space are the most proximal motor of development, and engagement with families is a lever of that quality, not an appendix of messaging (OECD, 2021). A digest extracts signals, orders them and pushes them to a phone. It can produce a notice. It does not produce dialogue.

The problem is sharpened by a professional fact that product sheets omit. In the nationally representative survey of public-school pre-K teachers in the United States, 82% used family-communication platforms (ClassDojo, Brightwheel and equivalents) and 75% did so daily or weekly; 84% agreed or strongly agreed that ed tech “could be helpful” for communicating with families (Berne, Doss, and Shapiro, 2025). That describes channel saturation. It does not describe pedagogical dialogue. In TALIS Starting Strong 2024, large shares of setting leaders report frequent informal communication with parents, and in many systems staff spend more than two hours a week in exchanges with families (OECD, 2025). That time belongs to adults who talk to one another. A model that compresses it into a paragraph does not inherit it.

This article does not recycle axes already treated in this series. The object is not parental mediation of children’s AI use, nor pedagogical documentation as an act of looking, nor emotion chatbots, nor a school-readiness score. The question is one of pedagogical category: what counts as family–school communication when an early childhood setting “does AI” on the school→family channel. The contributions are three: to reconstruct the state of the art that separates notices, digests and chatbots from pedagogical communication; to examine three families of cases; and to offer four tests for deciding when a kindergarten may claim that it communicates, and not merely that it notifies, condenses or generates.

2. State of the art: from notice to dialogue among responsible adults

Five strata that the “AI for families” market usually mixes need to be kept apart. The first is logistical communication: schedules, absences, health, menus, permissions. The second is visibility: photos, videos and fragments of life that show the child in the classroom. The third is pedagogical documentation: situated interpretation of what the child does and learns, which this article distinguishes from the automatic digest and does not redo as its object. The fourth is pedagogical family–school communication: dialogue among responsible adults—educator and mother, father or guardian—about the experience, care and learning of a particular child, in a time and a context both recognise. The fifth is generation: models that draft, condense, answer or translate those strata and claim to stand in for the fourth.

In the partnership stratum, NAEYC (2020) requires reciprocal associations: the educator takes responsibility for respectful relationships; seeks frequent two-way communication; and shares with the family knowledge of the particular child. Informing is not reciprocal. “Telling families about their child is a one-way form of communication”; two-way forms give families occasions to collaborate and to talk about concerns, goals and dreams (Steen, 2022). Hsu and Chen (2023), with 368 parents of young children in Taiwan, show that digital platforms are associated with parental involvement, teacher–child interactions and online communication, and that those factors explain 59.4% of the variance in family–school partnerships; the highest coefficient is online communication (0.473). Status: a finding of parental perception of platforms, not that a digest replaces dialogue. Inference: the channel can mediate partnership; mediation is not dialogue.

In the alignment stratum, Jensen, Dufur, Jarvis, and Pribesh (2025) analyse parents and teachers of 2,968 German kindergarten children aged 4–5 (NEPS). Differences in perception of the child’s desire for knowledge are negatively associated with all five developmental metrics examined; differences in talkativeness, confidence, good-naturedness and understanding are negatively associated with at least one outcome. Status: a finding that adult–adult disagreement about the child is not innocuous. Inference: a feed that “tells the day” without educator and family agreeing on who that child is does not repair the perceptual gap. It may widen it: each adult reads the digest from their own portrait.

In the channel stratum, León-Nabal, Zhang-Yu, and Lalueza (2021) document, in a classroom of 17 children and 15 families aged 3–6 in Barcelona, that the app—activated as two-way during the pandemic—is initiated mainly by the school and that its principal aim is to show classroom activities; families initiate, above all, logistics. Chen and Rivera-Vernazza (2023) observe ClassDojo in a private preschool: the platform can promote proactive involvement and partnership, but limited functions are used and digital communication has limits that face-to-face talk does not resolve. DiGiacomo, Greenhalgh, and Barriage (2021) recall that ClassDojo is not only a channel: it turns behaviour into points, stores them and shares them with families; it is datafication of classroom life, not dialogue about learning. Urbina, Ferrer-Ribot, and Villatoro Moral (2025) locate the problem, for ages 0–6, in digital tools as mediation of school–family communication. The object of this article is not that mediation in general: it is the leap to declaring pedagogical communication when a model generates the send.

In the rights stratum, the Recommendation on the Ethics of AI requires human supervision, proportionality and particular attention where children are concerned (UNESCO, 2021). UNICEF (2021) requires prioritising the best interests of the child and supporting development. The 2022 ethical guidelines and the U.S. Department of Education report agree on not substituting professional judgement or the teacher (European Commission, 2022; U.S. Department of Education, 2023). Miao and Holmes (2023) set an age threshold for independent conversations with generative platforms and require pedagogical validation. Inference: the three-year-old is not the user of a notice chatbot. The child is the subject of a conversation among adults that the chatbot does not sustain. Regulation (EU) 2024/1689 treats as high-risk, in Annex III, AI systems intended to evaluate learning outcomes or determine educational level; a day digest is not, in itself, that evaluation, but a summary that “evaluates the day” and is filed as a record approaches that threshold (European Union, 2024). Parental consent, when the summary implies the child’s data, authorises a processing operation. It does not authorise a pedagogical category.

3. Review method

A critical narrative review was conducted, not a meta-analysis. The purpose was not to estimate a homogeneous effect size among a neighbourhood app, a commercial platform and a generative digest, but to articulate an argument of pedagogical category with verified sources. Inclusion criteria: (a) 2021–2026; (b) school–family communication in early childhood education, notice apps, parental chatbots, automatic day digests or AI-enabled portals in ECEC; (c) relevance to kindergarten, preschool, CENDI or ages 0–6; (d) peer-reviewed journal, DOI or report from UNESCO, OECD, UNICEF, the European Union, the European Commission, NAEYC, RAND or a ministry; (e) a verifiable DOI or publisher page. Axes already used in this series were excluded: play and genAI scaffolding; UNESCO in-service teacher competence (Miao and Cukurova, 2024); parental mediation of children’s AI use; inclusion; adaptive tutors; student AI literacy; privacy as sole axis; multilingualism; teacher wellbeing; integrity; pedagogical documentation as object; low connectivity; initial teacher education; leadership; SEL and emotion chatbots; and preschool–primary transition or readiness scores.

The search was run on 25 August 2026 on DOI pages, Springer, Elsevier, Frontiers, MDPI, Taylor & Francis, RAND, OECD iLibrary, UNESDOC, UNICEF, EUR-Lex, ERIC, PubMed/PMC and NAEYC. Each source was checked against at least one of those pages. The corpus was organised into notice apps and platforms; automatic generation of messages, chatbots and digests; and evidence of everyday dialogue among responsible adults.

The analysis distinguished three enunciative statuses. Empirical finding: what was observed or measured in the sample. Conceptual or normative frame: what a framework, guide or regulation prescribes. Pedagogical inference: the translation to kindergartens, preschools, CENDI and infant schools, marked as such. The limits are those of a narrative review (section 9).

4. Case 1. A notice app that is “already two-way” is not pedagogical dialogue

León-Nabal, Zhang-Yu, and Lalueza (2021) publish in Frontiers in Psychology the study that best illustrates the first artefact. In a public school in Barcelona, in a working-class neighbourhood, with pupils mostly Roma or descendants of African, Asian or Latin American migration, they follow through telematic ethnography (September 2020–January 2021) the early-childhood team’s meetings and interaction with families via a communication app (Dinantia), activated as two-way during lockdown. They observe a classroom of 17 children, 15 families. Group messages—at least one every Friday, with text, images or video—report the week’s activities. Individual messages initiated by the school include mood, logistics and health. Those initiated by families number 21 in November and December: 11 logistical (schedules, absences), five thanks for photos, two health, two home activity and one mood. The school initiates most conversations. The main aim of group messages is to show what the child does in the classroom. Only six of the 15 families read every weekly message; two read none. Families do not, in general, use the app to share household activities, preferences or knowledge. The formal register of some bulletins—insecticides, mathematical concepts in block play, non-sexist toys—reproduces the asymmetry of hegemonic school culture. The coordinator states that “this year there is less bonding” and that “written messages are very impersonal” (León-Nabal et al., 2021).

Status of the evidence. Empirical finding on a digital channel at ages 3–6: when technical two-wayness is switched on, traffic remains, in substance, school→family, centred on displaying the classroom and on logistics. It is not a finding about generative AI. It is not a finding that “more messages” produce pedagogical communication. Pedagogical inference, marked as such: the gesture a kindergarten copies when it “puts an AI notice app in place” is exactly this, amplified. A notification channel is taken, a model is added to draft the weekly report, and communication is declared. What exists is a more fluent showcase. Platform two-wayness is not dialogical two-wayness. NAEYC (2020) asks that families be a source of information about the child and that educator and family share their knowledge. An app that does not receive funds of knowledge from the home does not meet that standard, even if the reply button is green.

Chen and Rivera-Vernazza (2023) saturate the portrait from ClassDojo. In a private centre in the northeastern United States they interviewed one teacher and three mothers of three-year-olds in the same classroom, and triangulated with platform artefacts. Four themes: modes of digital communication, the nature of that communication, limitations, and ClassDojo. In ClassDojo, three subthemes: proactive parental involvement, partnership building, and use of limited functions. Status: a qualitative finding that a commercial platform can mediate involvement and partnership in a middle-class classroom, and that only a fraction of what it promises is used. It is not a finding that ClassDojo is pedagogical communication. It is not a finding about automatic generation. Inference: the product sells “bringing every family into the classroom.” What Chen and Rivera-Vernazza observe is a subset of functions—messages, photos—used by a small number of adults. DiGiacomo et al. (2021), in a southeastern U.S. state, recall the other side of the same platform: points, avatars, behaviour turned into data shareable with the family. Inference: if what arrives at home is a point tally or an album, the setting has produced visibility or discipline. It has not produced dialogue about the child.

Berne, Doss, and Shapiro (2025) scale the phenomenon. With 1,586 public-school pre-K teachers, nationally representative, they find that family-communication platforms are the most used teacher-facing ed-tech type: 82% use, 75% daily or weekly. Eighty-four percent agree or strongly agree that ed tech could help communicate with families; among those who use it “more than half the time” or “every time” for that purpose, 90% agree or strongly agree, against 46% of those who never use it for that. Status: a finding of saturation and of perceived usefulness, not of dialogue quality or reciprocal partnership. Inference: the kindergarten that “already has an app” has solved the medium, not the category. The channel matters (Hsu and Chen, 2023). The digest does not, by that fact, inherit the weight of communication.

5. Case 2. Drafting the message or condensing the day is not talking with the family

The second artefact is generation: a model that writes the note, answers “what did they do today” or compresses the day into a paragraph. Berne et al. (2025) offer the closest, and the most sober, datum. Twenty-nine percent of public-school pre-K teachers used generative AI for job-related purposes in 2024–2025; 9% once a week or more. In the focus groups, one teacher said they used ChatGPT so that messages to families would “come across as less direct and more friendly.” Status: a finding of incipient use and of one case of tone-drafting. It is not a finding that the generated message improves partnership. It is not a finding of a day digest validated in ECEC. Pedagogical inference, marked as such: the gesture a setting copies when it “puts AI in the family portal” is this. The note the educator would have written is taken, a model is asked to soften it, and communication is declared. What exists is stylistics. A friendly tone is not dialogue. The family receives a voice that is no longer, entirely, that of the adult who was with the child.

That displacement of voice is the core of the second case. Miao and Holmes (2023) require pedagogical validation of generative AI and an age threshold for independent conversations with platforms. European Commission (2022) and U.S. Department of Education (2023) require not substituting the teacher. Inference: if the chatbot’s user is the family, not the child, the age threshold is not “met” by delegation. It is respected by not passing off a generated answer as the conversation the educator should sustain. UNESCO (2021) and UNICEF (2021) require human supervision and the best interests of the child. A digest sent without a responsible adult reading, correcting and situating it does not have that supervision. Consent to process photos and routines authorises the datum. It does not authorise calling a paragraph that nobody dialogued pedagogical communication.

Two objects the market mixes need distinguishing here. Pedagogical documentation is situated interpretation of what the child does: an adult’s gaze, made common, a hypothesis of learning. White, Rooney, Gunn, and Nuttall (2021) and Nuttall, Rooney, Gunn, and White (2023) have shown how digital documentation platforms push tagging, tracking and “completeness” of the record, and how that pressure is not, in itself, looking. That object was treated in this series as documentation. The automatic digest of the day is another operation: it condenses check-in, meals, nap, photos and, sometimes, a generated sentence, and pushes it to the family as “what happened today.” It does not interpret. It packages. Inference: a setting cannot invoke the documentation tradition to legitimate a digest. Documentation, when it is documentation, requires the adult. The digest spares the adult.

A third displacement clarifies the limit of the parental chatbot. Entenberg, Mizrahi, Walker, Aghakhani, Mostovoy, Carre, Marshall, Dosovitsky, Benfica, Rousseau, Lin, and Bunge (2023) publish a controlled trial of a 15-minute micro-chatbot that teaches 170 parents of children aged 2–11 to use positive attention and praise: they engage and recall the skills; there is no significant change in behaviour or self-efficacy at 24 hours. Status: a finding about a parenting chatbot, not a school–family channel. Inference: the market already sells automatic conversations to parents. That product is not pedagogical communication with the kindergarten. A “what did they do today” bot inherits that logic: the family talks to a system that was not in the classroom.

6. Case 3. What the kindergarten does when there is communication: everyday dialogue among adults

Fujisawa, Nozaki, Naoi, Shikishima, and Akabayashi (2025) publish in Early Childhood Research Quarterly the study that best anchors, in this corpus, what pedagogical communication actually is. With data on 239 children (141 boys, 98 girls; mean age 4.95 years, SD 0.99) in formal childcare, drawn from the Japan Child Panel Survey–Preschool Survey, they examine the role of everyday communication between parents and educators. That communication is positively associated with parental self-efficacy; self-efficacy is negatively associated with the child’s problem behaviour and with a conflictive parent–child relationship. Mediation analysis shows that self-efficacy mediates the association between parent–educator communication and those outcomes. Socioeconomic status moderates the communication–self-efficacy link: higher-SES parents benefit more. The finding the thesis needs is in the negative clause: the child’s everyday experiences at the centre were not associated with parental self-efficacy or with the other outcomes (Fujisawa et al., 2025). Human contact between adults was.

Status of the evidence. Empirical finding that, in Japanese childcare centres, what is associated with parental confidence and, through it, with the child’s behaviour and the relationship, is everyday communication with the educator, not the account of “what the child lived today.” It is not a finding about AI. It is not a finding that a digest reproduces that effect. Pedagogical inference, marked as such: this is the contrast an early childhood setting cannot evade. The automatic digest of the day promises exactly what Fujisawa et al. measure and find null: informing about the day’s experiences. The communication that counts is the one that occurs among adults, often in the exchange notebook and the doorway conversation characteristic of those centres. A model that writes “today painted, slept, ate well” substitutes the null object for the effective object. It declares communication. It delivers a feed.

Jensen et al. (2025) saturate the portrait from alignment. In 2,968 kindergarten children, parent–teacher disagreement about desire for knowledge is associated with worse developmental metrics in all domains examined. Inference: to communicate is not to emit. It is to bring two adult readings of the same child closer. A chatbot that answers the family without having spoken with the educator does not bring readings closer. It preserves the gap and gives it the format of a reply. TALIS Starting Strong 2024 shows, in most participating systems, that leaders report frequent informal communication with families—conversations about the child’s activities—and that home visits are rare; in many countries staff spend more than two hours a week in exchanges with parents (OECD, 2025). That time is the process quality that Starting Strong VI locates in interactions (OECD, 2021). Inference: if the setting “saves” those hours with a digest, it has not gained pedagogical efficiency. It has cut the mesosystem.

NAEYC (2020) closes the normative contrast: the educator seeks two-way communication; recognises that forms differ; uses drop-off and pick-up conversation, the conference and reciprocal technology; involves the family as a source of information about the child; shares knowledge of that particular child. Steen (2022) translates: telling the family about the child is one-way; dialoguing is two-way. The global report on early childhood care and education locates the right to a strong foundation in environments, educators and families that sustain learning and wellbeing, not in a sending indicator (UNESCO and UNICEF, 2024). Inference: evidence of pedagogical communication is verified in situated dialogue, in perceptual alignment and in the self-efficacy that is born of adult-to-adult dealing. If the “evidence” is a notification log, a notice chatbot or a generated summary, the setting has done product management, not partnership pedagogy.

7. Inferential frame: four tests for claiming communication, not a feed

The frame that follows is this article’s pedagogical inference, anchored in the cases and in the verified instruments. It is not a new international standard. It distinguishes four tests. If a kindergarten, preschool, CENDI or infant school does not pass them, it cannot declare that a notice chatbot, an automatic digest of the day or an AI-enabled portal constitutes pedagogical family–school communication.

7.1. Test of situated dialogue among responsible adults, not of the generated feed. Fujisawa et al. (2025) show that everyday parent–educator communication counts, not the account of the day’s experiences. NAEYC (2020) locates communication in conversations between educator and family. León-Nabal et al. (2021) show a channel that, even when two-way, does not produce that dialogue. Inference: evidence of communication is verified in two adults responsible for the child talking, listening and situating what they say in a shared context. If the “evidence” is a digest, a bot log or a generated weekly report, the setting has sent, not dialogued.

7.2. Test of two-wayness, not of the technical reply. NAEYC (2020) and Steen (2022) distinguish informing from reciprocal. León-Nabal et al. (2021) show families who do not send household knowledge and traffic dominated by the school. Chen and Rivera-Vernazza (2023) show limited functions. Inference: a reply button is not pedagogical two-wayness. Two-way means the family is a producer of knowledge about the child, not only a reader of a feed. A chatbot that “answers 24/7” can increase one-wayness: it gives the impression of interlocution and removes the interlocutor.

7.3. Test of pedagogical content and alignment, not of logistics or visibility. León-Nabal et al. (2021) quantify logistics, photos and thanks. DiGiacomo et al. (2021) describe behaviour points. Jensen et al. (2025) show that disagreement about the child is associated with worse outcomes. Inference: a menu, a lice notice or an album can be useful. They are not pedagogical communication. Neither is a summary that “shows the day” without bringing the two adult readings closer. Pedagogical communication is about who that child is, what they are learning, how they are cared for and what it is wise to do together. A digest of routines does not enter that conversation.

7.4. Test of professional voice and human supervision, not of generated tone. Berne et al. (2025) document the use of ChatGPT to soften notes. Miao and Holmes (2023), European Commission (2022), U.S. Department of Education (2023), UNESCO (2021) and UNICEF (2021) require pedagogical validation, not substituting the teacher, human supervision and the best interests of the child. Inference: an early childhood setting cannot treat the family as the user of a synthetic voice speaking in the educator’s name. A friendly tone does not replace the judgement of the person who was with the child. Consent to process the summary’s data does not turn the paragraph into dialogue. In European territory, a system that evaluates the child’s learning with AI and communicates it as a verdict approaches Annex III high-risk (European Union, 2024). The everyday digest is not, by default, that evaluation. Presenting it as a “generated pedagogical report” does push it towards that.

The frame admits apps as a channel of logistics and visibility (León-Nabal et al., 2021; Chen and Rivera-Vernazza, 2023; Berne et al., 2025; Hsu and Chen, 2023); pedagogical documentation as an adult’s gaze, not as a digest (White et al., 2021; Nuttall et al., 2023); informal communication and weekly time with families as a process-quality practice (OECD, 2021, 2025); and everyday parent–educator communication as a measured association, not as a feed (Fujisawa et al., 2025; Jensen et al., 2025; NAEYC, 2020). It refuses to declare pedagogical communication on the strength of a notice chatbot, an automatic digest of the day, a generated “friendlier” message, a portal that counts sends as partnership, or the treatment of the family as user of an interlocutor who was not with the child (Berne et al., 2025; León-Nabal et al., 2021; Entenberg et al., 2023; European Union, 2024).

8. Discussion

Three tensions organise the discussion. The first is between notifying and communicating. It is a finding that 82% of public-school pre-K teachers use family platforms and that 75% do so daily or weekly (Berne et al., 2025); that a two-way app at ages 3–6 remains, in substance, a school showcase of activities and logistics (León-Nabal et al., 2021); and that ClassDojo can mediate involvement and, at the same time, be used in a limited way or as a behaviour tally (Chen and Rivera-Vernazza, 2023; DiGiacomo et al., 2021). It is a frame that pedagogical communication is two-way, frequent and reciprocal (NAEYC, 2020). It is not a finding that saturating the channel produces the dialogue Fujisawa et al. (2025) associate with parental self-efficacy. The policy of the four artefacts—app, chatbot, digest, portal—measures what engineering knows how to measure (sends, reads, response time) and declares what only situated dialogue would authorise.

The second is between the account of the day and dealing among adults. Fujisawa et al. (2025) separate, with uncommon clarity, the child’s everyday experiences at the centre—not associated with self-efficacy or the other outcomes—from everyday parent–educator communication—yes associated, mediated by that self-efficacy. The automatic digest of the day bets on the first term. Partnership pedagogy lives in the second. Jensen et al. (2025) add that what is at stake is not only “being informed,” but coinciding, at least in part, on who the child is. A feed does not negotiate that coincidence. It delivers it. OECD (2021, 2025) locates informal conversation time and engagement with families in process quality. Inference: replacing those hours with a generated paragraph is not a process improvement. It is a subtraction disguised as personalisation.

The third is between the educator’s voice and the model’s voice. Berne et al. (2025) record the use of ChatGPT so that the note “sounds friendlier.” Entenberg et al. (2023) show chatbots that talk with parents about parenting, with skill learning and without behaviour change at 24 hours. Inference: generated friendliness is interlocution without an interlocutor. In early childhood, the family does not need a style: it needs the adult who saw the child. White et al. (2021) and Nuttall et al. (2023) warn, on the neighbouring ground of documentation, that the platform pushes tagging and completing the record. The digest performs the analogous operation towards home. Consenting to that send does not make it dialogue: it makes it data processing with a narrative.

9. Limits

This review is narrative. It does not apply PRISMA or estimate pooled effects. León-Nabal et al. (2021) observe a classroom in a pandemic; transfer to a generative chatbot is inference. Chen and Rivera-Vernazza (2023) is qualitative (one teacher and three mothers in a private centre). Berne et al. (2025) cover public-school pre-K in the United States; the 29% genAI use is broad job-related use, not specific to digests; the ChatGPT case comes from an unweighted focus group. Fujisawa et al. (2025) measure everyday communication in Japanese centres, not AI; the design is associative and SES moderates the benefit. Jensen et al. (2025) measure perceptual alignment, not a digital channel. Hsu and Chen (2023) are parental perceptions in Taiwan (65% university-educated). DiGiacomo et al. (2021) include students and principals. Entenberg et al. (2023) is a parenting chatbot. White et al. (2021) and Nuttall et al. (2023) are cited to distinguish documentation, not to redo that object. Urbina et al. (2025) are cited as a contemporary frame of digital tools at 0–6, without inflating sample findings not fixed here in the same detail. UNESCO, UNICEF, the OECD, the European Union, the European Commission, the U.S. Department of Education and NAEYC are prescriptive. No Latin American trials of a generative day digest in CENDI or kindergarten were located with the same degree of DOI. The inferences in section 7 are hypotheses of pedagogical category, not implementation evidence.

10. Conclusions

A notice chatbot, an automatic digest of the day or an AI-enabled portal does not constitute pedagogical family–school communication in an early childhood setting. The verified evidence does not authorise that declaration. A two-way app in a 3–6 classroom remains, in substance, a channel the school initiates to show activities and resolve logistics; that is digital mediation, not dialogue (León-Nabal et al., 2021). ClassDojo can mediate involvement and, at the same time, be used in a limited way or as a behaviour tally (Chen and Rivera-Vernazza, 2023; DiGiacomo et al., 2021). Eighty-two percent of public-school pre-K teachers already use family platforms; 29% use genAI, sometimes to soften the note; that is saturation of channel and of tone, not partnership (Berne et al., 2025). The child’s everyday experiences at the centre are not associated with parental self-efficacy; everyday communication with the educator is (Fujisawa et al., 2025). Parent–teacher disagreement about the child is associated with worse developmental metrics (Jensen et al., 2025). By contrast, when there is pedagogical communication, there is situated, two-way, reciprocal dialogue among responsible adults (NAEYC, 2020; Steen, 2022); there is weekly time for informal conversation (OECD, 2025); there is process quality in interactions, not in the log (OECD, 2021). Current rights require human supervision, the best interests of the child and not substituting the teacher (UNESCO, 2021; UNICEF, 2021; European Commission, 2022; U.S. Department of Education, 2023; Miao and Holmes, 2023).

Where the sources do not measure a kindergarten, this article does not assert it. Where they measure notice, visibility or tone generation, it does not translate them into pedagogical communication. Accompanying children from birth to six with their families is to sustain situated dialogue among the adults responsible for the child. The rest is a feed. It is not pedagogical communication, and it should not be presented as what it is not.

Editorial Laboratory of NEXTECH.IA / Ingeniero Mitre.

References

  1. Berne, J., Doss, C. J., & Shapiro, A. (2025). Pre-K teachers are optimistic about educational technology, though current use varies widely: Findings from the American Public School Pre-K Teacher Survey (RR-A4412-2). RAND Corporation. https://doi.org/10.7249/RRA4412-2
  2. Chen, J. J., & Rivera-Vernazza, D. E. (2023). Communicating digitally: Building preschool teacher-parent partnerships via digital technologies during COVID-19. Early Childhood Education Journal, 51(7), 1189–1203. https://doi.org/10.1007/s10643-022-01366-7
  3. DiGiacomo, D. K., Greenhalgh, S., & Barriage, S. (2021). How students and principals understand ClassDojo: Emerging insights. TechTrends, 66(2), 172–184. https://doi.org/10.1007/s11528-021-00640-6
  4. Entenberg, G. A., Mizrahi, S., Walker, H., Aghakhani, S., Mostovoy, K., Carre, N., Marshall, Z., Dosovitsky, G., Benfica, D., Rousseau, A., Lin, G., & Bunge, E. L. (2023). AI-based chatbot micro-intervention for parents: Meaningful engagement, learning, and efficacy. Frontiers in Psychiatry, 14, 1080770. https://doi.org/10.3389/fpsyt.2023.1080770
  5. European Commission. (2022). Ethical guidelines on the use of artificial intelligence (AI) and data in teaching and learning for educators. Publications Office of the European Union. https://doi.org/10.2766/153756
  6. Fujisawa, K. K., Nozaki, K., Naoi, M., Shikishima, C., & Akabayashi, H. (2025). Associations between daily parent–teacher communication, child’s problem behavior, and parent–child relationship mediated by parental self-efficacy. Early Childhood Research Quarterly, 72, 361–370. https://doi.org/10.1016/j.ecresq.2025.05.001
  7. Hsu, P.-C., & Chen, R.-S. (2023). Analyzing the mechanisms by which digital platforms influence family-school partnerships among parents of young children. Sustainability, 15(24), 16708. https://doi.org/10.3390/su152416708
  8. Jensen, M., Dufur, M. J., Jarvis, J. A., & Pribesh, S. L. (2025). Bridging the gap: The role of parent–teacher perception in child developmental outcomes. Children, 12(9), 1260. https://doi.org/10.3390/children12091260
  9. León-Nabal, B., Zhang-Yu, C., & Lalueza, J. L. (2021). Uses of digital mediation in the school-families relationship during the COVID-19 pandemic. Frontiers in Psychology, 12, 687400. https://doi.org/10.3389/fpsyg.2021.687400
  10. Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO. https://doi.org/10.54675/EWZM9535
  11. National Association for the Education of Young Children. (2020). Developmentally appropriate practice (DAP) position statement. https://www.naeyc.org/resources/position-statements/dap/contents
  12. Nuttall, J., Rooney, T., Gunn, A. C., & White, E. J. (2023). The impact of digital documentation platforms on early childhood educators’ work in Australia and New Zealand. Technology, Pedagogy and Education, 32(2), 257–273. https://doi.org/10.1080/1475939X.2023.2177720
  13. OECD. (2021). Starting Strong VI: Supporting meaningful interactions in early childhood education and care. OECD Publishing. https://doi.org/10.1787/f47a06ae-en
  14. OECD. (2025). Results from TALIS Starting Strong 2024: Strengthening early childhood education and care. OECD Publishing. https://doi.org/10.1787/20af08c0-en
  15. Steen, B. F. (2022). Five Rs for promoting positive family engagement. Teaching Young Children. National Association for the Education of Young Children. https://www.naeyc.org/resources/pubs/tyc/winter2022/fiver-rs-family
  16. UNESCO. (2021). Recommendation on the ethics of artificial intelligence. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000381137
  17. UNESCO & UNICEF. (2024). Global report on early childhood care and education: The right to a strong foundation. UNESCO. https://doi.org/10.54675/FWQA2113
  18. UNICEF. (2021). Policy guidance on AI for children 2.0. UNICEF Innocenti. https://www.unicef.org/innocenti/reports/policy-guidance-ai-children
  19. European Union. (2024). Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence. Official Journal of the European Union, L 2024/1689. https://eur-lex.europa.eu/eli/reg/2024/1689/oj
  20. Urbina, S., Ferrer-Ribot, M., & Villatoro Moral, S. (2025). School-family communication in early childhood education through digital tools. International Journal of Early Childhood. https://doi.org/10.1007/s13158-025-00419-3
  21. U.S. Department of Education, Office of Educational Technology. (2023). Artificial intelligence and the future of teaching and learning: Insights and recommendations. https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf
  22. White, E. J., Rooney, T., Gunn, A. C., & Nuttall, J. (2021). Understanding how early childhood educators “see” learning through digitally cast eyes: Some preliminary concepts concerning the use of digital documentation platforms. Australasian Journal of Early Childhood, 46(1), 6–18. https://doi.org/10.1177/1836939120979066