Is It OK to Use AI to Make Children's Books? Our Honest Answer

We use AI to help make our books, and we would rather say so plainly than have a parent find out later and feel misled. But saying it plainly is the easy part. The real question is whether we should, and that is worth a full answer, because it is the question we sat with before we made the first book.

Some families draw a firm line: nothing their children read should be made with AI. We understand that line, and we respect it. If it is yours, then we are probably not your imprint, and we would rather tell you that than argue you out of a value you hold for good reasons.

Most of the parents who write to us are worried about something narrower. They are not against a tool on principle. They are against the slop: the lazy AI-written book already flooding the storefronts, and the older thing behind it, the Cocomelon churn built to hold a child's eyes and leave nothing behind. We hate both. The tools that made them possible can be pointed the other way, and that is what we are trying to do. So here is how a book is actually made here, and where a person is standing at each step.

What "no AI, ever" would actually cost

We are two people: one dad and his mother, who taught for thirty years. We wanted a particular kind of book for our own family, a real subject like Plutarch built all the way up, from a board book a toddler can hear to a prep book that readies a teenager for the original text. That book did not exist, and the reason was plain: no one was ever going to fund a niche subject built this fully by hand. So we started making it ourselves.

The choice a family faces, then, is not our books against a hand-made version of the same thing. No one has built that version, and we see no sign anyone will. The real choice is between our books, the classic retellings we grew up on and still love, and a plain encyclopedia entry. The tools are what let two people build the whole ladder instead of a single book. That is the trade we weighed, and it is why we said yes.

Where the people are

Every book is shaped by us before any AI touches it. We read the source, decide what the book is really about, and build its plan, scene by scene. Only then does the AI draft the prose and render the pictures, inside the plan we have already drawn, and we rework every page by hand until it is what we set out to make.

Then the part that decides whether a book is any good is ours. We check every claim against the original source. We put every draft through an independent read that hunts for anything weak, padded, or machine-sounding, and nothing that helped write a book is ever allowed to approve it. Karen and I make every final call, and we read each book cover to cover before it goes out. Nothing ships that we would not want on our own family's shelf. There is more on exactly how all of this works in How we use AI.

The test we would use

There is a simple way to know whether one of our books reads like a living thing or like filler, and it does not require trusting us at all. Read it aloud, and have your child tell it back to you. If it is hollow, they will have nothing to say, and nothing we claim here should change your mind. If it is alive, they will have plenty, and then how a book was made matters a good deal less than what it is.

We did not pick up these tools to make books faster and cheaper. We picked them up to make the books we could not find, and to get a child to the real thing sooner. Anything less than that, and they would not be worth using.