“Educational” AI YouTube videos accused of teaching kids to play in traffic & eat toxic food

by

Experts in children’s media are warning about AI-generated videos on YouTube that claim to be educational but actually include concerning or damaging content.

Recent reports from Futurism and The 74 show a surge in AI-created videos for children on platforms like YouTube. These videos often look like educational content kids already trust, but they contain significant mistakes.

Sometimes, these errors are seriously unsafe. For example, one video showed children walking in a road with oncoming traffic, as if it were a normal thing to do. Another video gave wrong information about traffic laws, saying “green means right” instead of “go.”

A new AI-created sing-along about the 50 states has some glaring errors. The video shows incorrect visuals paired with audio, and mispronounces state names – for example, calling Rhode Island “Ribio Island,” Connecticut “Conmecticut,” Oklahoma “Oklolodia,” and Louisiana “Louggisslia” – as if these were the correct names.

Videos showing babies and characters eating potentially dangerous foods are causing worry. Some clips feature babies eating whole grapes – a common choking hazard – and honey, which can make infants sick with botulism. Others show people eating raw elderberries, which are poisonous if not cooked first.

Health experts concerned over YouTube’s “educational” AI slop

Experts caution that these materials aren’t without risk. Young children learn best through repeating things and seeing them, so wrong information can easily get stuck in their minds. This is especially true when the information is presented in a colorful and appealing way that seems reliable.

Carla Engelbrecht, a former employee of Sesame Street and PBS Kids, expressed growing alarm at the amount of content she was finding online, calling it “downright dangerous.” She warned that the more she discovered, the more concerned she became.

University of Chicago professor Dana Suskind described the spread of inaccurate information to young children as “toddler AI misinformation at an industrial scale,” warning that it poses significant risks to their developing brains.

The problem stems from the growing trend of “AI slop” – videos quickly made using automated tools with little human supervision. These tools can create scripts, images, and voiceovers in just minutes, letting channels pump out a lot of content focused on getting as many views as possible.

Kapwing, a video editing company, recently discovered that around 20% of videos on YouTube are now created using artificial intelligence, and these AI-generated videos are collectively bringing in millions of dollars for their creators.

Experts advise parents not to automatically trust a video just because it’s labeled “educational.” Instead, they suggest carefully watching what their children are viewing, rather than simply going by the video’s picture or title.

YouTube is now actively removing poor-quality content created by AI. They’ve already taken down channels that had amassed billions of views and are asking viewers to identify content that seems artificially generated and unoriginal. Videos flagged by users are then being removed.

YouTube is experimenting with a new feature that aims to prevent misleading thumbnail images, often used to attract clicks.

Growing numbers of health organizations are expressing concern that AI-generated content could hinder a baby’s development, potentially blurring the line between reality and fiction as they grow up.

Read More

2026-03-25 17:50