The use of artificial intelligence services in generation of illustrations for children’s literature
This work is devoted to the interaction of artificial intelligence in creating illustrations for children’s literature. Rapid advances in artificial intelligence (AI) have transformed creative industries, shifting the designer’s role from that of sole creator to that of“manager” who controls AI-based optimization. The relevance of this study lies in the critical role that illustrations play in a child’s development – acting as a “bridge” between reader and text for shaping personality and emotional intelligence – and the consequent need for high- quality, ethically sound visual content in educational materials.The research methodology uses comparative analysis of traditional design and AI-generated illustrations. The practical component included creating a series of black-and-white outline drawings for elementary school textbooks (grades 1–2) and comparing illustrations created by a professional designer using Adobe Illustrator with images created by two AI tools: Large Language Model (LLM), Nano Banana 2, and the visual generation platform Leonardo.ai. To ensure consistency, “engineering prompting” was used, which allowed for a structured request algorithm of six blocks (The Goal, Negative Constraints, Style, Subject, Geometry, and Complexity Control) to limit “creative freedom” AI and compliance with strict technical requirements. The results indicate that AI services can create contoured illustrations, but retain significant differences between human and machine performance. The Nano Banana 2 tool provided more “cartoon” and detailed results suitable for children, while Leonardo.ai offered cleaner lines but sometimes logical inconsistencies in overlapping lines. The common flaws created by AI included asymmetric forms, unnatural anatomical structures, and visual “noise”. By contrast, the human-designed illustrations demonstrated compositional logic, precise symmetry, and stylistic integrity. The study concludes that while generative AI can serve as an “assistant/helper” for creating initial sketches and optimizing workflow, final products now require extensive manual refinement by a professional designer. The results confirm that human and artificial intelligence collaboration is possible, but still depends on the accuracy of requests (prompts) and human control.
Authors
- Daria Yerokhina
- Нонна Кулішова
Institutions
- Kharkiv National University of Radio Electronics (UA)
Publication Details
- Journal
- Innovations in Publishing Printing and Multimedia Technologies
- Published
- 2026-11-20
- DOI
- https://doi.org/10.59476/ilpmt2026.v2i2.803
- Primary Topic
- Digital Media and Visual Art
- Type
- article
- Field-Weighted Citation Impact
- 0.00