AI-Generated Kids’ Videos Are Flooding YouTube

YouTube is rapidly filling up with AI-generated videos targeting young children, sparking growing concern among parents and child-safety advocates. While these clips often mimic familiar nursery rhymes or books, they frequently contain bizarre imagery, unnatural voices, and even factual errors. In some cases, videos intended to teach language skills have mispronounced basic words, creating confusion for toddlers still developing speech. It has led to a problem where children are increasingly shaped by poorly regulated, machine-made content.

The issue reached headlines after comedian and father Alex Pearlman shared a viral post describing his toddler’s encounter with an AI-generated book adaptation on YouTube Kids. The video imitated an illustrated vocabulary book but delivered distorted narration, mislabeling, and mispronouncing basic objects.

Here's the AI generated adaptation of Ms. Rachel's book.

For many parents, it symbolized a new challenge of how to ensure the reliability of media consumed by children who often cannot distinguish real from artificial.

Parent groups have since raised alarms, warning that these videos exploit trust in digital platforms branded as “safe” for minors. Critics argue that labeling and filtering systems on services like YouTube Kids have not kept pace with generative AI, leaving families exposed to content that slips through moderation tools. Pressure is now mounting on both tech companies and regulators to address the rapid rise of machine-generated children’s programming.

Industry experts in digital safety say the problem stems from the accessibility of AI production tools, which have made creating children’s content nearly effortless. Scott Kollins, chief medical officer at Aura, a family online safety company, explained that “AI makes it possible to generate these clips in bulk, purely to monetize views, with little regard for accuracy or impact.”

Dr. Natalie Bidnick Andreas, assistant professor of communication studies at the University of Texas at Austin, has warned in her research that content generated this way frequently contains bizarre or inaccurate details masked by familiar characters. She and other media literacy scholars caution that beyond miseducation, such clips may also incorporate unsettling or even age-inappropriate elements, all while appearing outwardly child-friendly.