Child-development research, applied to every video
Researchers and pediatric organizations have spent years studying what screen time does for young children and what it does to them. That work now gives parents more than a time limit: peer-reviewed findings and institutional guidance on which kinds of content to steer clear of, and what healthy screen habits look like.
We built that research into a written standard and paired it with KidSight AI that lets us apply the standard to every video in the library, at a scale no parent could manage alone, so the research is already behind the choice when you make it.
1
We ingest the video itself
Every frame, sound and word in the video. Never the description or someone else’s review.
2
KidSight AI reads it
The AI behind our engine records what happens on screen and in the audio, frame by frame, bit by bit.
3
The rubric scores it
Our written standard turns those readings into a verdict for each age, consistently for every video.
4
It earns a place in the library
Our team publishes it, and it reaches your child only if it fits their age and clears the safety checks.
KidSight AI
What KidSight AI reads in a video
KidSight AI powers our engine, and it starts with the video itself. The footage and the full audio track go to the model together, and it reads the every frame and the soundtrack side by side: what happens on screen, what’s said, the music and sound effects, and any words that appear on screen. We never rely on other people's reviews and judgement - we make our own judgement based on what our AI sees.
What KidSight AI records
What it teaches, and how it goes about teaching it
How it moves and how it sounds
Who is in it
What it asks a child to do
A set of safety readings
KidSight AI isn’t trained to have opinions about children’s video. It applies the same written criteria, drawn from published guidance, to every video, in the same order, every time.
The rubric
The reading and the rubric are kept separate
KidSight AI · reads
Analyzes what’s in the video
Reports what it finds against written definitions, including how well the video suits each age. It doesn’t decide whether a video belongs in the library.
The rubric · scores
Turns readings into a verdict for each age
The rubric is our written standard, with criteria drawn from published child-development guidance. It weighs KidSight AI’s readings the same way, in the same order, for every video, and produces a separate verdict for each age.
Fit for the age
Weighted most heavily: how well the video suits the age, and its pace, sensory intensity, vocabulary and emotional tone.
Teaching
Early reading, numbers, shapes and colors, science, problem solving and everyday skills, and whether a video actually teaches them or only features them.
Social and emotional
Naming and managing feelings, getting along, kindness and conflict worked through, weighted differently at each age.
Craft
Production quality, storytelling, performance and voice.
Safety
Frightening content, aggression, mature themes and crude language rule a video out at an age. Ads and product-pushing are flagged.
Why the two are kept apart
If one system both described a video and judged it, a change in our standards would be impossible to tell apart from a change in the AI. Keeping them apart means we can update the rubric as more studies are published, and we can run the whole library through it again, and see exactly what moved.
Checked before it goes live
Every version of the rubric is checked before it goes live. If anything in it is incomplete, or points at a reading that doesn’t exist, the release stops. Nothing is allowed to quietly count as zero because of a mistake in how it was written down.
WHY AGE MATTERS
How the answer changes with age
Pace shows it most clearly. KidSight AI places every video on a four-point scale of pace, and measures how much is happening on screen and in the audio alongside it.
Pace is never scored as good or bad on its own. The rubric weights it differently depending on how old the child is.
Pace, weighted by age
Calm peaks at two, then falls away. Fast does the reverse.
How much a video’s pace counts toward its fit at each age, on a scale of 0 to 1. These are weights, not scores.
The research
Where the standard comes from
The rubric’s criteria are drawn from published guidance by leading child-development organizations, each of which has put years of study into how children use screens and what helps or harms them. The social and emotional criteria follow the CASEL framework for social and emotional learning.
No one can pay to be in the library.
No creator has a relationship with us, and none has any say in what we publish or what we say about a video.
American Academy of Pediatrics
Family media guidance
Age fit, pace, watching together, vocabulary, curiosity and science, critical thinking and everyday skills.
ZERO TO THREE
The 5 C’s of media use
The social and emotional criteria, sensory intensity, emotional tone and watching together.
NAEYC and the Fred Rogers Center
Technology position statement
Early reading, numbers, shapes and colors, problem solving and carrying learning into real life.
Harvard Center on the Developing Child
Serve and return
Warmth: whether the people on screen respond to a child the way a caring adult would.
The library
How a video gets into the library
Chosen by our team
Every video in the library has been analyzed by KidSight AI and then deliberately published by a person on our team. Nothing publishes without a human in the loop.
Matched to your child’s age
A video appears for your child only if it fits their age and clears every safety check for that age.
Lots of options to choose
We make sure there is intentional variety in your lineup so you can discover new shows and characters that resonate better with your kids. But don't worry, we still have your favorites.
