YouTube Kids Under Pressure: Generative AI Worries Parents

A Collective Letter for the Protection of Children

More than 230 organizations and experts recently sent a letter to Sundar Pichai, CEO of Google, and Neal Mohan, CEO of YouTube. This initiative, led by the Fairplay association, aims to ban AI-generated content on YouTube Kids. The signatories, including researchers and pediatricians, emphasize the need to protect children by preventing the recommendation of such videos to those under 18 and strengthening parental controls to block this content.

This move comes shortly after Google's announcement of a $1 million investment in Animaj, a studio specializing in creating AI-generated content for young audiences, including infants. The authors of the letter use this example to support their call for stricter regulation.

The Dangers of Automated Content for Children

The signatories of the letter describe the consequences of automatic recommendations after watching popular children's videos, such as "Cocomelon." They estimate that nearly 40% of suggested videos are AI-generated and of poor quality, a phenomenon they refer to as "AI Slop." Some channels that broadcast this content generate annual revenues exceeding $4.25 million. These videos often present inconsistencies, such as letters displayed on screen that do not match the sounds, or songs with distorted words. Children, unable to read, risk internalizing these errors without any possibility of correction.

Dana Suskind, a professor at the University of Chicago, emphasizes that this content can harm language development and comprehension in young children. Kathy Hirsh-Pasek, a psychology professor at Temple University, notes that some videos claim to be educational but contain incorrect information.

YouTube has established guidelines for creators of content aimed at children, stressing the importance of age-appropriateness, encouragement of learning, and a coherent structure. However, the signatories believe that AI-generated videos do not adhere to these principles, displaying discordant elements and exploiting popular keywords without providing real educational value.

Rachel Franz, director of the Young Children Thrive Offline program at Fairplay, explains that these videos captivate children beyond their attention span, making it difficult for parents to control their children's exposure to this content.

Recommendations for a Stricter Framework

Experts propose six essential modifications to improve the situation:

• Clear labeling of all AI-generated content.
• An option in parental controls to block these videos, even if a child searches for them.
• Removal of automated recommendations for minors regarding AI content.
• Cessation of Google's investments in the creation of AI-generated videos for children.

YouTube responded by stating that the company already imposes labeling rules for AI-generated content that imitates real people or scenes. However, cartoon-style animations, which dominate children's videos, are not subject to these requirements. The platform also offers parents the ability to block channels and report videos.

The signatories remind that a California jury recently found YouTube negligent in protecting young users during a trial concerning social media addiction. They view this decision as a call for firmer action to protect young audiences.

Furthermore, some creators use generative tools to produce educational and structured videos. The authors of the letter insist that these formats should not mask the proliferation of low-quality automated content. They urge YouTube to adjust its rules and recommendations to limit children's exposure to this content until reliable criteria can distinguish truly suitable productions from mere automated creations.