1. Introduction and problem

In the 3–6 age range—daycare, preschool, CENDI, early childhood school—a package of three artifacts has been installed that purports to stand in for artistic creativity. The first is an app that “completes” the child’s drawing: a contour recognition or assisted fill system that finishes the figure, corrects the stroke, or proposes the “well done” version and declares that the classroom “already does art with AI.” The second is an image generator from a prompt: a text-to-image model, activated by the adult or by a child’s phrase, that delivers a polished print and exhibits it as evidence of the group’s creativity. The third is a model that scores “creativity”: an automated classifier or rubric that assigns a score to visual products and converts that number into proof that the center “develops creative thinking.” All three are visible and cheap in coordination time. They allow the center to display that it “already does artistic creativity with artificial intelligence.” The leap—from obtaining a completed image, a generated print, or a score to claiming that there is process-based artistic creativity—is not authorized either by the evidence on painting or AIGC in childhood or by what NAEYC and the process art tradition understand when there is material, bodily, and meaning-based exploration in the daily life of the classroom.

The thesis of this article is restrictive. An app that completes the child’s drawing, an image generator from an adult or child prompt, or a model that scores “creativity” does not constitute artistic creativity in early childhood education. At this stage, artistic creativity is a process of exploration with materials, body, and meaning—process art versus product-focused art; teacher mediation that listens to and sustains the child’s intention; creative imagination in the Zone of Proximal Creative Development—not a generated image output nor a creativity score. NAEYC (n.d.), in its editorial resource on the development of creativity, distinguishes process art from product art: the value lies in doing, choosing, trying out, and making meaning, not in a standardized final product. That does not authorize translating “there is a generated image” or “there is a creative score” as “there is artistic creativity.” It authorizes asking what was measured or what was displayed: often, technical feasibility, engagement with an interface, or improvement of composition and color under active reference, not the material and bodily exploration that a four-year-old initiates when they smear, knead, tear, or combine without a closed model.

The problem is aggravated by four category confusions that product sheets do not mention and that this article separates with rigor. First: artistic creativity is not free play—already addressed in another piece in this series. Second: it is not STEM nor educational robotics. Third: it is not pedagogical documentation—portfolios or panels that record what occurred. Fourth: it is not social-emotional education (SEL) as a program of emotional competencies. This work does not recycle articles on free play, STEM, documentation, or SEL. The question is one of artistic creativity: what counts as process art and mediation of imagination when an early childhood center “does AI and creativity.” Confusing the product—completed drawing, generated print, score—with the process is the error this work names. Wang, Zhang, Wang, and Zheng (2025) warn of cognitive homogenization through standardized interfaces and recommend process-oriented evaluation. Zhang, Binti Halili, and Zainuddin (2026) point out, in a review with SWOT analysis of twenty-one studies, the explicit threat of potential reduction of children’s creativity. Gong, Ma, and Wang (2026) show that passive reception of AIGC inhibits originality in idea generation, while active reference during the process can improve design expression. Pedagogical inference, marked as such: the market for “AI for artistic creativity in early childhood” inherits that evaluative distribution and converts it into a promise of creativity. It converts the generated image and the score into the whole of the artistic craft.

What the three artifacts share is an economy of visibility that favors coordination over craft. A completed contour, a polished print, and a numeric score travel well in newsletters, parent meetings, and accreditation folders. They signal that the center is “innovating” without requiring observers to watch a four-year-old knead clay, negotiate color with a peer, or revise a figure because the story changed mid-stroke. Pedagogical inference, marked as such: that visibility is not neutral. It rewards what can be photographed, exported, or averaged—and it quietly displaces what process art actually demands: time, mess, hesitation, and an adult who does not rush the product toward closure.

The contributions are three: reconstruct the state of the art that separates artistic creativity (process art; material and bodily exploration; teacher mediation; creative imagination) from completing drawings, generating images, and scoring creativity; examine three families of empirical cases; and offer four tests to decide when a daycare center can claim that there is artistic creativity, and not merely an app that completes, a generator, or a scorer.

2. State of the art: from artistic creativity to the artifact on display

It is useful to separate four strata that the market for “AI for artistic creativity in early childhood” usually mixes. The first is the construct of artistic creativity from ages 3 to 6 as process art, material and bodily exploration, and mediated creative imagination—not product art nor generative output—(NAEYC, n.d.; NAEYC, 2022; Loizou and Loizou, 2022; OECD, 2021, 2023). The second is the teacher craft that cultivates it—listening, sustaining intention, Solve/Reflect/Share/Connect, human-in-the-loop—(Loizou and Loizou, 2022; Berson and Berson, 2024; Nikolopoulou, 2025). The third is the evidence on painting with AI, GenAI in preschool, AIGC as visual inspiration, digital training systems, and mappings of AI in ECE (Wang et al., 2025; Zhang et al., 2026; Gong et al., 2026; Chen, Lin, and Chien, 2022; Zeng et al., 2025; Chen, 2024; Su and Yang, 2022; Ljungcrantz, 2026). The fourth is the framework of rights, systems, and developmentally appropriate practice, which treats the 3–6 child as a subject of artistic exploration, not as a user of a generation or scoring pipeline (UNESCO, 2021; Miao and Holmes, 2023; European Commission, 2022; U.S. Department of Education, 2023).

In the stratum of the construct, NAEYC (n.d.) establishes process art versus product-focused art: the process of exploring materials, making decisions, and making meaning matters more than a uniform final product. NAEYC (2022) requires developmentally appropriate practice: the 3–6 child is a subject of meaningful experiences, not of standardized outputs. Loizou and Loizou (2022), from the cultural-historical activity theory, situate the teacher as mediator of creative play and propose the Zone of Proximal Creative Development with Solve, Reflect, Share, and Connect strategies; they distinguish process and product orientation. Status: empirical-conceptual framework of mediation, not of automatic generation. OECD (2021) anchors ECEC quality in meaningful everyday interactions; OECD (2023) situates digitalization in uses with pedagogical meaning, not in screens that substitute for exploration. Inference: a dashboard of “scored creativity” or a gallery of generated images does not cover, by existing, process art nor creative imagination.

In the stratum of craft, Berson and Berson (2024) document CultureCraft: intersection of AI, historical imagery, and early childhood creativity with human-in-the-loop; AI does not substitute for the teacher nor for human interaction in preschool. Nikolopoulou (2025) warns that GenAI can limit autonomy and creativity if it is prescriptive, and calls for child-centered sociocultural mediation. Status: finding and framework that creativity and human interaction remain on the side of the adult guarantor. It is not a finding that a prompt generator reproduces that craft. Inference: an app that completes the drawing does not listen to the child’s intention. An adult who sustains the process does.

In the stratum of artifacts, Wang et al. (2025) systematically review, with PRISMA, twenty empirical articles on AI-based painting technology and children’s creative thinking: they find risks of cognitive homogenization through standardized interfaces and recommend process-oriented evaluation. Zhang et al. (2026) synthesize twenty-one studies of GenAI in preschool education with SWOT analysis and identify an explicit threat of potential reduction of children’s creativity, along with problems of reliability and age appropriateness. Gong et al. (2026) contrast passive reception of AIGC—which inhibits originality in idea generation—with active reference during the process—which improves design expression in composition and color; they conclude that AIGC is auxiliary, not a substitute. Chen, Lin, and Chien (2022) test a digital training system with AI (contour recognition, color matching): finding of technical feasibility, not that “completing” the drawing equals artistic creativity. Zeng, Rahim, and Xu (2025), with N = 60 in third grade of primary school, report engagement and self-efficacy with the MAI-SGD tool: transfer marked—not a 3–6 sample; does not authorize declaring daycare creativity. Chen (2024), Su and Yang (2022), and Ljungcrantz (2026) map AI in ECE without equating it to daycare process art.

In the stratum of systems, UNESCO (2021) requires human supervision and AI ethics. Miao and Holmes (2023) set pedagogical validation and age thresholds for generative AI. The European Commission (2022) and the U.S. Department of Education (2023) agree on not substituting professional judgment. Inference: a four-year-old is not the passive recipient of a prompt-generated image nor the object of a creativity scorer. They are the protagonist of a material and bodily exploration that an adult guarantor listens to, names, and sustains without closing the product. Pedagogical inference, marked as such: when policy documents insist on human-in-the-loop and developmentally appropriate practice, they are naming the same adult function that Loizou and Loizou (2022) describe as mediation of creative play—not the function of an interface that completes, generates, or scores on the child’s behalf.

3. Review method

A critical narrative review was conducted, not a primary meta-analysis. The purpose was not to estimate a homogeneous effect size of GenAI on creativity, but to articulate a pedagogical category argument with verified sources. Inclusion criteria: (a) 2021–2026, with transfer explicitly marked when the sample is not 3–6; (b) artistic creativity, process/product art, GenAI/AIGC, painting with AI, teacher mediation, material exploration, or AI in early childhood education; (c) relevance for daycare, preschool, CENDI, or ages 3–6; (d) peer-reviewed journal, DOI, or report from NAEYC, UNESCO, OECD, European Commission, or department of education; (e) verifiable DOI or editorial page. Excluded as central object were the axes already used in this series—free play, STEM/robotics, pedagogical documentation, SEL, executive functions, generative tutoring, gaps, privacy, UDL, numeracy, formative assessment, family–school, continuing education, CLASS interaction quality—although some appear only as category limits.

The search was executed on August 28, 2026 (slot 05:02 America/Mexico_City) on DOI pages, Frontiers, Wiley, Springer, OECD iLibrary, UNESDOC, JAIR, NAEYC, CEDTECH, iJIM, and editorial sites. Each source was verified against at least one of those pages. Empirical finding, conceptual or normative framework, and pedagogical inference marked as such were distinguished. Priority was given to the distinction artistic creativity / free play / STEM / documentation / SEL, and to the caution of Wang et al. (2025) and Zhang et al. (2026) regarding homogenization and potential reduction of creativity. Literature whose exact authorship could not be confirmed on an editorial page at the time of verification was not cited.

4. Case 1. Completing the drawing, generating the image, or scoring “creativity” does not constitute artistic creativity

Chen, Lin, and Chien (2022) publish in Frontiers in Psychology the system that best names the first artifact when it is presented as creativity: a digital training environment with AI-assisted learning for color perception in drawing, with contour recognition and chromatic matching. Status of the evidence. Empirical finding of technical feasibility: the system can assist digital learning of color and contour. It is not a finding that installing an app that “completes” or corrects the drawing develops or guarantees process-based artistic creativity in ages 3–6, nor that material and bodily exploration are covered. The data the market does not cite is the ceiling: technical assistance was demonstrated, not equivalence with process art. Pedagogical inference, marked as such: this is the gesture a daycare center copies when it “does creativity with AI that finishes the drawing.” A stroke is traced, completion occurs, the clean figure is displayed. What is present is an assistance artifact. Artistic creativity, in NAEYC (n.d.) and in Loizou and Loizou (2022), requires exploration and mediation of intention, not a contour “well resolved” by the model. When a teacher hands a tablet to a child and the app “corrects” the circle into a perfect shape, the classroom may gain a displayable product; it does not automatically gain the trial-and-error, bodily engagement, or narrative intention that process art requires. The child’s partial figure may have been the beginning of a story the adult never heard because the model closed the form first.

Gong, Ma, and Wang (2026) saturate the portrait of the second artifact—the generator and AIGC visual inspiration—from the contrast between passive reception and active reference. They find that passive reception of AIGC content inhibits originality in idea generation; that active reference during the process can improve design expression in composition and color; and that AIGC must remain auxiliary, not a substitute. Status: finding that mode of use matters and that passivity harms originality. It is not a finding that an adult or child prompt that generates a polished print constitutes the child’s creative imagination. Inference, marked as such: delivering the generated image as the “group’s artwork” is product art amplified by GenAI. A group project in which the adult types “happy dinosaur in the jungle” and hangs the resulting image on the wall documents the model’s capacity and the adult’s curatorial choice; it does not, by itself, document whether each child explored material, revised intention, or narrated meaning during the activity. Zhang et al. (2026), in a systematic review with SWOT of twenty-one studies on GenAI in preschool, name an explicit threat: potential reduction of children’s creativity, along with doubts about reliability and age appropriateness. Status: synthesis of threat to the very construct the market promises to enhance. Inference: a CENDI that exhibits Midjourney-like galleries as proof of creativity has done generation. It has not demonstrated process art.

Wang, Zhang, Wang, and Zheng (2025) close the third artifact—the scorer and evaluation of “creative thinking” via painting with AI—with a PRISMA review of twenty empirical articles. They document the impact of AI-based painting technologies on children’s creative thinking and, critically, risks of cognitive homogenization through standardized interfaces; they recommend process-oriented evaluation. Status: synthesis finding on homogenization and mandate for process evaluation. It is not a finding that a model that scores creativity measures or produces process art. Zeng et al. (2025) report, with N = 60 in third grade of primary school, engagement and self-efficacy with MAI-SGD: transfer marked—not 3–6; does not authorize declaring daycare artistic creativity nor validating a creativity scorer in early childhood. Chen (2024), Su and Yang (2022), and Ljungcrantz (2026) confirm growth of AI in ECE without evidence that completing, generating, or scoring substitutes for the craft. Inference: a daycare center that delivers “AI-generated creativity” or “sufficient creative score” has done output or measurement. Artistic creativity is verified in whether the child explored materials and body with meaning and whether the adult sustained that intention. It is not verified in the generated PNG nor in the scorer’s number.

5. Case 2. What the preschool does when there is artistic creativity: process art, exploration, and mediation

NAEYC (n.d.) defines process art versus product-focused art as a current institutional editorial framework: the value lies in exploring, deciding, trying out, and making meaning with materials, not in reproducing a closed model. Status: verifiable institutional conceptual framework on the NAEYC editorial page. It is not an AI finding. Pedagogical inference, marked as such: this is the object an early childhood center can call artistic creativity. Exploration is not a prompt: it is body, matter, and meaning. An app that completes the drawing without reading intention can produce a “correct” figure and, at the same time, truncate the trial. A generator can deliver a beautiful print and not have captured whether the child chose, mixed, or narrated their own process. In a process-art classroom, the adult notices when a child returns to the same color three times, when they abandon a figure to start another, or when they explain that the smudge “is rain.” Those moments are the empirical surface of artistic creativity in ages 3–6. They are not recoverable from a scorer’s JSON nor from a generator log unless the adult has already observed and named them—and if the adult has, the digital artifact was never the source of the creativity claim.

Loizou and Loizou (2022) saturate the craft from cultural-historical theory: the teacher mediates creative play; they propose the Zone of Proximal Creative Development and the Solve, Reflect, Share, and Connect strategies; they distinguish emphasis on process and on product. Status: empirical-conceptual framework of mediation. Inference: a daycare center cannot treat “we already generate art with AI” as equivalent to having mediated creative imagination. Berson and Berson (2024) provide the human-in-the-loop bridge with CultureCraft: creativity and human interaction in preschool; AI does not substitute for the teacher. Status: finding/design case of responsible practice. Inference: noticing intention is a condition of mediation; a generator does not listen on behalf of the teacher.

Nikolopoulou (2025) warns that child-centered GenAI integration must balance promises and risks: prescriptive GenAI can limit autonomy and creativity; sociocultural mediation is needed. Status: synthesis framework oriented to ECE. NAEYC (2022) requires developmentally appropriate practice: the 3–6 child is a subject of experience, not of pipeline. OECD (2021) anchors quality in meaningful daily interactions; OECD (2023) requires digitalization that empowers without substituting. Gong et al. (2026) reinforce, from the AIGC side, that only active reference during the process—not passive reception—can assist composition and color, and even then as auxiliary. Inference: when there is artistic creativity in early childhood, there is process art (material and bodily exploration with meaning), teacher mediation that listens and sustains, and creative imagination in the proximal zone; not an average of scores nor a gallery of outputs. Wang et al. (2025) recommend process-oriented evaluation precisely because standardized interfaces homogenize. Inference: process art, exploration, and mediation are verified at the art table, the yard, and the workshop—when the adult sustains intention without closing the product—not in the generator log nor in the scorer’s JSON.

6. Case 3. Artistic creativity is not free play, nor STEM, nor documentation, nor SEL

The first category boundary is free play. Free play—already addressed in another piece in this series—can empirically overlap with exploration; it does not, by itself, constitute the construct of artistic creativity as mediated process art in visual arts and creative expression. Status: pedagogical inference of series distinction, marked as such. This article does not convert free play into a central object: it names it to prohibit the equivalence. An app that completes drawings is not free play either: it is algorithmic assistance to the product.

The second boundary is STEM/robotics. This article does not convert artistic creativity into computational thinking, coding, or robotic assembly. Chen (2024), Su and Yang (2022), and Ljungcrantz (2026) are retained for mappings of AI in ECE, not for STEM. Inference: when a product promises “develops STEM creativity because it generates images” or “trains design because it completes the drawing,” it has crossed to another construct or collapsed art with engineering. Artistic creativity can dialogue with other languages; it is not identified with them.

The third boundary is pedagogical documentation. A digital panel, a portfolio, or an anecdotal record can preserve traces of what occurred; they do not, by themselves, constitute the process art that NAEYC (n.d.) anchors in doing. Status: pedagogical inference anchored in the process/product framework, marked as such. This article does not convert documentation into a central object: it names it to prohibit the equivalence. An image generator is not documentation of meaning either: it is synthetic production. Confusing “we have works on screen” with “there was artistic exploration” is the category error the market exploits.

The fourth boundary is SEL. This work does not recycle a SEL article: it does not identify artistic creativity with global socioemotional competencies. OECD (2021, 2023) and NAEYC (2022) are retained for interactions, digitalization, and appropriate practice, not for SEL programs. Inference: when a product promises “improves creative self-esteem because it scores creativity” or “develops emotional expression because it generates an avatar,” it has crossed to another construct. Zhang et al. (2026) and Nikolopoulou (2025) warn of reduction or limitation of creativity under GenAI; that finding/framework does not authorize selling GenAI as artistic SEL. Miao and Holmes (2023) set age limits and validation. Inference: the only use of AI coherent with ages 3–6 remains on the side of the adult who mediates exploration—as support for active, not passive, inspiration (Gong et al., 2026), or for teacher formation—subject to pedagogical validation. It does not enter as an app that completes the drawing, a generator that substitutes for the process, nor a scorer that declares creativity fulfilled.

7. Inferential framework: four tests to claim artistic creativity, not an artifact

The framework that follows is pedagogical inference from this article, anchored in the cases and in the verified instruments. It is not a new international standard. It distinguishes four tests. If a daycare center, preschool, CENDI, or early childhood school does not pass them, it cannot declare that an app that completes the drawing, an image generator, or a model that scores creativity constitute artistic creativity.

7.1. Test of process art and material-bodily exploration, not of completed or generated output. NAEYC (n.d.) defines process art versus product art. Chen, Lin, and Chien (2022) demonstrate feasibility of digital assistance to contour and color. Gong et al. (2026) show that passive reception of AIGC inhibits originality. Inference: evidence of artistic creativity is verified in whether the child explored materials, body, and meaning. If the center’s “evidence” is the drawing completed by the app or the generator’s print, the center has done assisted or generated product art, not process art. Pedagogical inference, marked as such: coordinators should ask not “what did the screen show?” but “what did the child do with matter and body before, during, and after the interface?” Without that question, process art collapses into product art with a digital finish.

7.2. Test of teacher mediation and the child’s intention, not of the prompt that substitutes. Loizou and Loizou (2022) document mediation and the Zone of Proximal Creative Development. Berson and Berson (2024) document human-in-the-loop: AI does not substitute for the teacher. Nikolopoulou (2025) warns of prescriptive GenAI. Inference: artistic creativity requires an adult who listens to and sustains intention. A prompt that “improves” or “finishes” the work without that listening does not demonstrate mediation. The Solve, Reflect, Share, and Connect strategies name an adult posture—problem-solving alongside the child, reflecting on choices, sharing discoveries, connecting to prior experience—that no completion algorithm performs. Pedagogical inference, marked as such: if the only “mediation” recorded is an adult typing a prompt, the center has documented operator behavior, not teacher craft.

7.3. Test of process-oriented evaluation, not of the creativity score. Wang et al. (2025) recommend process evaluation in the face of homogenization through standardized interfaces. Zhang et al. (2026) warn of potential reduction of children’s creativity with GenAI. Zeng et al. (2025) measure engagement and self-efficacy in primary school (transfer marked), not daycare creativity. Inference: “creativity = 8.7” can elevate a number without elevating exploration or imagination. A score is not process art.

7.4. Test of category distinction and professional judgment, not of the product catalog. Artistic creativity ≠ free play / STEM / documentation / SEL. UNESCO (2021), Miao and Holmes (2023), the European Commission (2022), the U.S. Department of Education (2023), NAEYC (2022), and OECD (2021, 2023) require human supervision, pedagogical validation, developmentally appropriate practice, and meaningful interactions. Inference: a center cannot treat the young child as a user of a generation or scoring pipeline. Artistic creativity is not fulfilled by generating a better image. It is fulfilled by exploring with materials and body, mediating intention, and sustaining creative imagination with adults who exercise the craft.

The framework admits the digital when it is subordinated to teacher mediation and to active—not passive—reference during the process (Gong et al., 2026; Berson and Berson, 2024; Nikolopoulou, 2025). It rejects declaring artistic creativity through an app that completes, a prompt generator, or a creativity scorer (Chen, Lin, and Chien, 2022; Wang et al., 2025; Zhang et al., 2026; Zeng et al., 2025).

8. Discussion

Three tensions organize the discussion. The first is between displaying a generation or scoring artifact and exercising process-based artistic creativity. It is a finding that AI systems assist contour and color (Chen, Lin, and Chien, 2022); that passive AIGC inhibits originality while active reference can assist design (Gong et al., 2026); that reviews warn of homogenization and potential reduction of creativity (Wang et al., 2025; Zhang et al., 2026). It is a framework that artistic creativity is played out in process art and mediation (NAEYC, n.d.; Loizou and Loizou, 2022; Berson and Berson, 2024). It is not a finding that completing, generating, or scoring produce the craft that Loizou and Loizou (2022) describe as creative mediation. The three artifacts deliver what engineering knows how to display and declare what only mediated exploration would authorize.

The second is between product art amplified by GenAI and creative imagination. Gong et al. (2026) and Nikolopoulou (2025) agree on the risk of passivity or prescription. Inference: insisting that the daycare center “already has artistic creativity” because the model generated beautiful prints or completed drawings is an inverted pedagogy. An output correlated with visual engagement is taken to stand in for process art. The young child becomes a provider of initial stroke or prompt; the adult becomes the operator of the generator. Pedagogical inference, marked as such: inverted pedagogy also inverts accountability. When the print is beautiful, the center celebrates the tool; when the print is generic or the score is low, the center blames the child or the prompt—not the absence of material exploration or mediation. Process art restores accountability to the adult who sustains intention, not to the pipeline that delivers polish.

The third is between creativity score and process evaluation. Wang et al. (2025) discourage interfaces that homogenize and call for process-oriented evaluation. Zeng et al. (2025) do not authorize, by transfer, extrapolating primary school engagement to 3–6 creativity. Inference: the only use of AI coherent with ages 3–6 remains on the side of the adult who mediates exploration, subject to pedagogical validation and to the guidelines of UNESCO, the European Commission, and the U.S. Department of Education. An autonomous scorer because the system needs a product is not that. A generator that substitutes for doing because coordination needs “pretty” photos for the display case is not either.

Additional pedagogical inference, marked as such: in contexts where OECD (2021, 2023) and NAEYC (2022) require meaningful daily interactions, an “AI art” session that occupies twenty minutes of screen time without return to material, body, or narrative does not satisfy the criterion of artistic exploration—even if the final print is aesthetically successful. Wang et al. (2025) and Zhang et al. (2026) converge on the risk of homogenization and reduction of creativity when the interface standardizes production. Inference: a center that wants to document artistic creativity must link each digital artifact to a trace of exploration observed by the adult—material choices, gestures, verbalizations—and not to the model’s visual output alone. Berson and Berson (2024) illustrate this with CultureCraft: AI enters as historical reference under teacher mediation, not as closure of the child’s drawing. Nikolopoulou (2025) adds that prescriptive GenAI limits autonomy; the coherent path remains child-centered sociocultural mediation before any claim of “creativity guaranteed by AI.”

Additional pedagogical inference, marked as such: the series of artifacts—app that completes, generator, and scorer—shares an economy of visibility: clean figure, print, number. NAEYC (n.d., 2022) and Loizou and Loizou (2022) require verifying creativity in process and mediation. Zhang et al. (2026) document threat of reduction of creativity with GenAI. Chen (2024), Su and Yang (2022), and Ljungcrantz (2026) show growth of AI in ECE without authorizing the equivalence. UNESCO (2021), Miao and Holmes (2023), the European Commission (2022), and the U.S. Department of Education (2023) subordinate AI to professional judgment: the adult remains guarantor of process art.

9. Limits

This review is narrative. It does not apply its own PRISMA nor estimate combined primary effects—although it cites the PRISMA review of Wang et al. (2025). Chen, Lin, and Chien (2022) demonstrate feasibility of digital assistance, not daycare process art. Gong et al. (2026) contrast passive reception and active reference of AIGC; they do not equate AIGC to 3–6 artistic creativity. Wang et al. (2025) synthesize twenty studies with risk of homogenization; they do not validate a commercial scorer. Zhang et al. (2026) review twenty-one GenAI studies in preschool with SWOT; the threat of reduction of creativity is from synthesis, not from a single Latin American RCT. Zeng et al. (2025) are N = 60 in third grade: transfer marked. Berson and Berson (2024) are a CultureCraft case with human-in-the-loop, not a trial of an app that completes drawings. Loizou and Loizou (2022) are a framework of creative mediation, not of GenAI. Nikolopoulou (2025) is a GenAI integration framework. Chen (2024), Su and Yang (2022), and Ljungcrantz (2026) map AI in ECE, not process art of visual arts. NAEYC, UNESCO, and OECD are framework sources. No Latin American trials of AI comparing adult-mediated process art versus app that completes, autonomous generator, or creativity scorer in CENDI were located. The inferences in section 7 are pedagogical category hypotheses, not implementation evidence.

In transfer to Latin American contexts, this article maintains caution: the cited designs do not authorize a daycare center to declare artistic creativity for having purchased a generator or a scorer. If technology is introduced, it enters on the side of the adult and as active reference—never as a substitute for doing—(Gong et al., 2026; Berson and Berson, 2024). Wang et al. (2025) and Zhang et al. (2026) brake scored creativity dashboards. Artistic creativity is not exhausted in free play nor identified with STEM, documentation, or SEL: it is verified when the adult sustains material, bodily, and meaning-based exploration.

10. Conclusions

An app that completes the child’s drawing, an image generator from an adult or child prompt, or a model that scores “creativity” do not constitute artistic creativity in an early childhood education center. The verified evidence does not authorize that declaration. Chen, Lin, and Chien (2022) demonstrate feasibility of digital training with AI for contour and color, not equivalence with process art. Gong et al. (2026) show that passive reception of AIGC inhibits originality and that only active reference can assist design as auxiliary, not substitute. Wang et al. (2025), in twenty studies, warn of cognitive homogenization and call for process-oriented evaluation. Zhang et al. (2026), in twenty-one GenAI studies in preschool, point to potential reduction of children’s creativity. Zeng et al. (2025) measure engagement in primary school (N = 60) with transfer marked. In contrast, when there is artistic creativity in early childhood, there is process art and mediation: process/product distinction (NAEYC, n.d.); Zone of Proximal Creative Development and Solve, Reflect, Share, Connect strategies (Loizou and Loizou, 2022); human-in-the-loop without substituting for the teacher (Berson and Berson, 2024); alert against prescriptive GenAI (Nikolopoulou, 2025); developmentally appropriate practice and meaningful interactions (NAEYC, 2022; OECD, 2021, 2023). Artistic creativity is distinguished from free play, from STEM/robotics, from pedagogical documentation, and from SEL. Current guidelines require human supervision, pedagogical validation, and not substituting professional judgment (UNESCO, 2021; Miao and Holmes, 2023; European Commission, 2022; U.S. Department of Education, 2023; Chen, 2024; Su and Yang, 2022; Ljungcrantz, 2026).

Where the sources do not measure a daycare center, this article does not claim they do. Where they measure completing drawings, generating images, or scoring creativity, it does not translate them into process-based artistic creativity. Accompanying three- to six-year-old children in artistic creativity is exercising process art: exploration with materials, body, and meaning; mediation that listens to and sustains intention; creative imagination. The rest is an app that completes, a prompt generator, and an algorithmic creativity score. It is not artistic creativity in early childhood education, and it must not be presented as what it is not.

Laboratorio Editorial de NEXTECH.IA / Ingeniero Mitre.

References

  1. Berson, I. R., y Berson, M. J. (2024). Fragments of the past: The intersection of AI, historical imagery, and early childhood creativity. Future in Educational Research, 2(4), 403–421. https://doi.org/10.1002/fer3.46
  2. Chen, J. J. (2024). A scoping study on AI affordances in early childhood education. Journal of Artificial Intelligence Research, 81, 701–740. https://doi.org/10.1613/jair.1.16882
  3. Chen, S.-Y., Lin, P.-H., y Chien, W.-C. (2022). Children’s digital art ability training system based on AI-assisted learning: A case study of drawing color perception. Frontiers in Psychology, 13, 823078. https://doi.org/10.3389/fpsyg.2022.823078
  4. 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
  5. Gong, W., Ma, J., y Wang, Z. (2026). Application of AIGC-assisted visual inspiration in early childhood art education. World Journal of Educational Studies. https://doi.org/10.61784/wjes3121
  6. Ljungcrantz, L. (2026). The interaction of AI and early childhood education. A state-of-the-art review 2020–2024. Early Childhood Education Journal, 54, 3565–3581. https://doi.org/10.1007/s10643-025-02079-3
  7. Loizou, E., y Loizou, E. K. (2022). Creative play and the role of the teacher through the cultural-historical activity theory framework. International Journal of Early Years Education, 30(3), 527–541. https://doi.org/10.1080/09669760.2022.2065248
  8. Miao, F., y Holmes, W. (2023). Guidance for generative AI in education and research. UNESCO. https://doi.org/10.54675/EWZM9535
  9. NAEYC. (2022). Developmentally appropriate practice in early childhood programs serving children from birth through age 8 (4.ª ed.). NAEYC. https://www.naeyc.org/resources/pubs/books/dap-fourth-edition
  10. NAEYC. (s. f.). Supporting the development of creativity. https://www.naeyc.org/our-work/families/supporting-development-creativity
  11. Nikolopoulou, K. (2025). Child-centered integration of generative AI in early learning: Balancing promises and challenges. AI, Brain and Child. https://doi.org/10.1007/s44436-025-00023-1
  12. OECD. (2021). Starting Strong VI: Supporting meaningful interactions in early childhood education and care. OECD Publishing. https://doi.org/10.1787/f47a06ae-en
  13. OECD. (2023). Empowering young children in the digital age (Starting Strong). OECD Publishing. https://doi.org/10.1787/50967622-en
  14. Su, J., y Yang, W. (2022). Artificial intelligence in early childhood education: A scoping review. Computers and Education: Artificial Intelligence, 3, 100049. https://doi.org/10.1016/j.caeai.2022.100049
  15. UNESCO. (2021). Recommendation on the ethics of artificial intelligence. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000381137
  16. U.S. Department of Education, Office of Educational Technology. (2023). Artificial intelligence and the future of teaching and learning: Insights and recommendations. U.S. Department of Education. https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf
  17. Wang, A., Zhang, Y., Wang, A., y Zheng, W. (2025). The impact of AI-based painting technology on children’s creative thinking. Frontiers in Psychology, 16, 1598210. https://doi.org/10.3389/fpsyg.2025.1598210
  18. Zeng, S., Rahim, N., y Xu, S. (2025). Integrating mobile AI in art education: A study on children’s engagement and self-efficacy. International Journal of Interactive Mobile Technologies (iJIM), 19(11), 112–142. https://doi.org/10.3991/ijim.v19i11.54847
  19. Zhang, Y., Binti Halili, S. H., y Zainuddin, Z. (2026). Generative AI in preschool education: A systematic review with SWOT analysis. Contemporary Educational Technology, 18(1), ep626. https://doi.org/10.30935/cedtech/17866