Child development specialists are calling on Google to ban videos created using artificial intelligence from being displayed to young audiences on YouTube and YouTube Kids. They express concerns over the low quality of AI-produced children’s content, arguing it’s detrimental to kids’ attention spans and their ability to distinguish between reality and fiction.
In a letter addressed to Google CEO Sundar Pichai and YouTube CEO Neal Mohan, over 200 experts from various advocacy groups and educational institutions cited what they view as a lack of educational value in many AI-generated videos. They are particularly worried about the proliferation of poor-quality children’s content being mass-produced by AI creators, which seems to prioritise profit over the well-being of young viewers.
The advocates labelled this type of content, sometimes referred to as “AI slop,” as harmful because it may impair children’s attention spans and replace necessary real-world interactions vital for their emotional and social growth.
“There is much we don’t know about the consequences of AI content for children,” the letter states. “YouTube is participating in this uncontrolled experiment by distributing AI-generated content without any research to support its benefits and disregarding child development principles that suggest it could be mostly harmful.”
Signatories of the letter include social psychologist Jonathan Haidt, author of “The Anxious Generation,” who has led efforts to combat the negative effects of social media on youth. Other supporters include child advocacy organisations like Fairplay and the National Alliance to Advance Adolescent Health, along with the American Federation of Teachers and numerous schools.
YouTube’s spokesperson, Boot Bullwinkle, responded by highlighting the company’s commitment to high standards for YouTube Kids, including limiting AI-generated content to a select number of quality channels. Bullwinkle emphasised that parents can also block certain channels and assured that YouTube strives for transparency with AI content, requiring creators to disclose any artificial elements.
The rise of AI-generated content on YouTube has become increasingly notable, particularly among creators targeting young children. Many have discovered that using AI can simplify and reduce the costs associated with video production. Some have even begun sharing methods to monetise this AI-assisted approach for generating content aimed at toddlers. In response, Bullwinkle noted that YouTube’s monetisation policies are designed to discourage the production of low-quality, spam-like content.
Mohan stated earlier this year that addressing the issue of “AI slop” and ensuring that YouTube remains a welcoming space for viewers are priorities for the company moving forward. Nevertheless, YouTube maintains that not all AI-generated content is of poor quality, asserting that, when executed correctly, it can be beneficial.
YouTube requires creators to mark content that has been altered or synthesised. However, advocates argue that these labels are often not comprehensible to the young viewers who are the intended audience for much of this so-called “AI slop.”
In March, Google announced an investment in Animaj, an AI animation studio aimed at producing children’s content for YouTube. While some within Google hailed this initiative as a progressive step towards enhancing offerings for young audiences, child safety advocates criticised the move, arguing that it targets infants and toddlers who should ideally have no screen exposure. They urged YouTube to cease all investments in generating AI-based videos for children.
The recent letter arrives amidst broader movements pressing for reform in how YouTube operates. In a related context, a jury trial earlier this year found Google and Meta Platforms Inc. liable for detrimental outcomes experienced by a young user, citing products designed to keep users engaged. Both companies announced plans to appeal, but advocates and lawmakers are now pushing for significant changes in YouTube’s operational practices, including its content algorithms.
