At Library For All, we strongly believe that children have the right to read texts that mirror their experiences and languages. Our programs deliver culturally-relevant books to children’s reading programs. We also strive to establish a culture of literacy, where the writing of a story, or planning and creating illustrations, is part of developing communities of engaged readers. In recent years, Generative AI has entered the publishing and education industries, with rapid uptake bringing some remarkable opportunities.
At the same time, for our mission to raise the voices of new and emerging storytellers, this technology brings fresh challenges and ethical dilemmas.
Collaborative content creation
Our regional collections are designed to be more than just sets of books; they are curated libraries, including a range of levels and themes, specifically suited to user groups. Wherever possible, we create original books by engaging communities in writers’ workshops to share their authentic stories. This approach means that children benefit from familiar vocabulary, local imagery and collective cultural wisdom as they improve their literacy. The authors get immense satisfaction from participating, and carry that excitement about writing, reading, and book creation back into their communities when they share their published books.
In our workshops, we demonstrate some of the key factors in writing a great story for children, such as a compelling opening and useful keywords, while leaning into local preferences for narrative structures and content. Some participants write long stories, some provide just a few key ideas, and our editorial team then shapes the story into book form, with careful attention to the balance of educationally appropriate grammar and authentic voice.
Learn more about our collections: Why Education
Sharing stories from a distance
In some projects, of course, such as our work in conflict zones or natural disaster response, it’s not possible for us to travel or provide online story sharing workshops. In these cases, we have the option to reach out via email or social media to invite local participants to write for us. This approach was essential during Covid, for example, and when building our library for Ukraine, which includes 50 original stories shared by local people at the beginning of the current conflict.
But in 2025, we’ve noticed a distinctive shift, with many of the stories coming to us via online discussions showing the telltale signs of AI generation. They are more formulaic, for example, and conform to rigid (usually US) English grammar conventions. They may include cultural content but have much less cultural ‘flavour’. We have found ourselves in a whole new world when it comes to selecting and editing community stories.
For example,
- Is a story AI generated by someone from a given culture still a culturally authentic story?
- Is it reasonable for us to remunerate these authors in the same way we do our other creative participants?
- Should we develop a firm policy rejecting AI story sharing. If so, how would we prove or enforce it?
- Are we being too rigid? Perhaps this in fact is giving a voice to authors who may not have otherwise considered sharing a story with us, particularly if they are living in a difficult or dangerous circumstances?
For now, our editorial decision making remains a case-by-case discussion, with a view to surveying community responses and preferences as books make their way to readers.
Explore our programs: Impact
Illustration for impact
Just like our stories, the illustrations in Library For All books serve several important functions. They help make books bright and appealing for young readers, they assist literacy by helping readers match words with images, and they support ‘visual literacy’ – that is, the understanding of how pictures represent ideas. This is a vital skill in today’s world where our children are growing up amidst the confusion of misinformation and fake news.
Generative AI tools can create pictures in seconds. So it’s no surprise that, as a non profit publisher, we are regularly asked why we wouldn’t use this technology to save costs and speed up our production. We find ourselves now asking new questions about what it means to be culturally safe in our illustrative work. There are also ramifications for clear storytelling.
One example comes from the Vanuatu workshops last year. We have a wonderful local story about a mother and child at home, who hear a frightening noise outside. They see a thief coming to take some bananas. Luckily their dogs come to the rescue and scare the thief away. Some of our editorial colleagues generously shared some generated images with us for potential inspiration in the book.
At this stage, our policy is not to use generated images for publication, but suggestions like these can be useful to guide our illustrators. In this case, however, we allowed our illustrator to interpret the story in their own way, and the results are significantly different.
Generative AI tools work on text prompts. So, prompts using words like ‘scared’ or ‘protect’ appear to have resulted in very literal representations. It is also difficult to get continuity of a character across generated images, so we see a different mother and child in each generated picture.
On the other hand, our illustrator has interpreted the text to show a more natural and nuanced version of the story, where the child is scared (but not utterly terrified) and the mother is protecting the child with kindness and trust in the dogs, not by hiding under a blanket. The artist also added the playful visual element of the dogs dreaming about bananas, which brings humour to the story for added reader appeal, and invites conversation around comprehension.
The challenge for responsible, educational storytelling here is that there is a difference between simply adding a picture that matches a set of words, and illustrating a story in a way that’s beautiful, relatable and engaging. In this case, our editorial team was satisfied that the artist’s version was the appropriate way to go. But over time, as Generative AI technology improves, perhaps things will change?
Library For All is a learning-focussed organisation, with a keen interest in ongoing awareness of best practice publishing and optimal ways to serve communities. But Generative AI has introduced some hitherto unimaginable new considerations for our publishing team when it comes to bringing culturally-relevant and educationally responsible books to children. For now, the key is ongoing dialogue with policy makers and reading communities alike to consider how we get the balance right.
A version of this content was originally shared in a presentation at the Australian School Library Association Conference 2025 by Dr Lara Cain Gray, Head of Acquisition (Publishing).
Follow our story: