Former YouTube Kids employees describe how the service embraced human curation over virality to address criticism it wasn't doing enough to protect children
Former employees say platform embraced human curation over virality to address criticism it wasn't doing enough to protect younger consumers Tweets: @yoreekoh , @kevinroose , and @martinhoffmann Tweets: Yoree Koh / @yoreekoh : As tech companies (ahem Instagram) think about creating kids' versions of their products they should look to YouTube Kids for lessons learned. The takeaway: it's really hard to build a safe and high quality platform for kids https://www.wsj.com/... Kevin Roose / @kevinroose : So much of the 2010s social media story boils down to tech executives' excessive faith in AI and lack of appreciation for human judgment in media. https://www.wsj.com/... https://twitter.com/... Martin Hoffmann / @martinhoffmann : „Holding on to young viewers involved a difficult reckoning at YouTube, the first tech giant to build a children's version of its main product. Ultimately it decided that humans, not machine-trained algorithms could best select content for children." https://www.wsj.com/...
Context & Ripple Effects
The arc here runs from scandal to quiet redesign. James Bridle's 2017 account of traumatizing videos reaching children through YouTube's recommendation system put algorithmic distribution for kids under scrutiny, and by mid-2019 critics were arguing YouTube should invest in curation rather than funnel everything into YouTube Kids, where the same problems would replicate (that Variety-era debate).
What changed, per the former employees, is that YouTube Kids accepted the editorial identity it once resisted: Bloomberg reported in late 2019 that YouTube had weighed screening all videos for under-8s but dropped the idea to avoid looking like a media company. Human curation winning out means that resistance collapsed — and the WSJ's framing lands just as Instagram runs its own campaign to win teen users, making YouTube Kids the cautionary case study other platforms building kids products are being told to study.
First-order effects
- YouTube Kids' product economics change at the root: hand-curated catalogs replace recommendation-driven growth, trading scale and engagement for a defensible answer to child-safety criticism.
- Instagram, mid-campaign to win back teens per internal documents, inherits a template with a known price tag — Koh's takeaway is that building a safe, high-quality kids platform is genuinely hard, which raises the bar before Meta ships one.
Second-order effects
- Any competitor building a kids version of a social product must now budget for human review as a permanent cost line, not a launch fix — curation headcount becomes table stakes where minors are the audience.
- The unmarked 'kidfluencer' sponsored-content problem sits outside curation's reach, so YouTube still faces the legal exposure Bloomberg flagged even with a cleaner catalog.
Third-order effects
- If the pattern holds, platforms serving children converge on a media-company model — editorial selection over virality — effectively splitting the industry into curated tiers for kids and algorithmic tiers for everyone else.
- That split hardens into a trust moat: with more than one in five videos YouTube suggests to new users reported as AI-generated slop, demonstrated human oversight becomes a competitive asset regulators and parents can price.
The trend: Consumer platforms are selectively abandoning pure algorithmic distribution in favor of human editorial control wherever child-safety and regulatory risk outweigh engagement gains.