My son lost it in the cereal aisle at two and a half. My phone was out of my pocket before I had a conscious thought about it. Not a decision, a reflex.
That reflex is the thing the American Academy of Pediatrics quietly built a framework around this year, and almost nobody covering the story picked up on it, because everyone was busy arguing about the clock.
The AAP stopped arguing about the hour
In January 2026 the AAP published "Digital Ecosystems, Children, and Adolescents: Policy Statement" in Pediatrics (2026;157(2):e2025075320, Munzer et al., Council on Communications and Media). The coverage was all about one thing: the rigid hourly limits are gone.
Here is what the statement actually says about limits, word for word:
> "Time limits might range from <1 hour/d for toddlers and preschoolers to 1 to 2 hours/d or more of entertainment (not school-related) media for school-aged children and teens. The most important considerations are high-quality content and prioritizing healthy activities (eg, sleep, play, physical activity, reading)."
Notice where time sits in that paragraph. It is the setup, not the point. The point is the sentence after it.
The replacement framework is five questions, which the AAP calls the 5 Cs: Child, Content, Calm, Crowding out, Communication. Four of those get plenty of attention. The third one is the one I would write on the fridge, and the AAP's own phrasing of it is six words:
> "Calm: Are children using media for calming?"
Not how long. Not what app. Whether the screen is the thing that ends the crying.
What the evidence on calming actually says
The statement is blunt about this, in a way policy documents usually are not:
> "For preschoolers, use of mobile devices for calming is associated with weaker emotion regulation skills and problematic media habits."
That sentence carries citations. I followed them.
The main one is Radesky, Kaciroti, Weeks, Schaller, and Miller, published in JAMA Pediatrics (2023;177(1):62-70). It followed 422 children with a mean age of 3.8 years across three waves over six months, with data collected between August 2018 and January 2020. Among boys, using devices for calming at wave two predicted higher emotional reactivity at wave three (β = 0.20; 95% CI, 0.10 to 0.30). Among children high in what researchers call surgency (high energy, impulsive, sensation seeking), the association ran in both directions: calming use predicted more reactivity, and more reactivity predicted more calming use.
A separate group found the same loop in a different country, in a different era. Fitzpatrick and colleagues, also in JAMA Pediatrics (published August 12, 2024), followed a community convenience sample of 315 preschoolers in Nova Scotia at ages 3.5 (2020), 4.5 (2021), and 5.5 (2022). A one standard deviation increase in tablet time at 3.5, which worked out to 1.22 hours a day, was associated with a 22 percent of a standard deviation increase in expressions of anger a year later (β = 0.22; 95% CI, 0.01 to 0.44; P = .04). Then it ran backward: a one standard deviation increase in anger at 4.5 was associated with a 22 percent of a standard deviation increase in tablet use at 5.5 (β = 0.22; 95% CI, 0.01 to 0.43; P = .04).
Two cohorts, one pre-pandemic and one squarely inside it, pointing the same direction. That is worth something.
It is also worth saying what these studies are not. Both are observational. Both are a few hundred children. The Nova Scotia authors flag their own limitations directly: a convenience sample, parent-reported tablet use, no measure of what content the children were watching, and a pandemic context they say could itself explain the instability they observed. These are associations across groups of children. They are not a mechanism proven in a lab, and they are not a statement about any individual kid.
The reason the loop makes sense
Strip the statistics off and the logic is almost boring.
Emotion regulation is a skill. Skills are built by doing the hard version of the thing, badly, over and over. A toddler learns to come down off a feeling by being inside the feeling with an adult nearby, and eventually coming down.
When the tablet appears, the feeling ends. Which is the entire point, and it works, and that is exactly the problem. The child stops being upset without having done any of the work of stopping being upset.
The AAP puts it in one word:
> "Using digital devices for soothing can displace opportunities for children to develop skills managing their emotions."
Displace. Not damage, not harm. Displace. The screen occupies the slot where the repetition would have gone.
That framing also explains why the usual defense does not apply here. On most screen questions, content quality is the strongest lever, and the AAP says so explicitly. On this one it is close to irrelevant. The best-designed, most carefully educational program in the world still ends the meltdown before the child has practiced ending it.
This is not an edge case
The 2025 Common Sense Census surveyed 1,578 parents of children age 8 and under through a probability-based panel, fielded August 5 to 29, 2024, with a margin of error of plus or minus 2.9 percent. It found that one in five families now use mobile devices to help manage their child's bedtime routines, mealtimes, and emotional regulation.
The same report found that 40 percent of children have their own tablet by age 2, and 58 percent by age 4.
So the device is present, it is the child's own, and in a fifth of homes it has an explicit job in the hardest moments of the day. Whatever this is, it is not a fringe behavior.
What actually changes at home
The AAP's clinical guidance here is one sentence, and it is a two-step, not a ban:
> "Pediatric providers can help families identify use of media for calming and offer alternative strategies."
Identify first. Most parents I talk to have no idea how often they do it, because the whole thing takes four seconds and lives below conscious thought. Mine did.
Then substitute. The statement suggests asking what is underneath the reach for the device, giving boredom as an example. In my house the honest answer was usually that I was tired and out of patience, which is a different problem than my son being tired and out of patience, and it needed a different fix.
The other thing I would flag from the Nova Scotia data is the direction nobody watches: harder kid, more tablet. If the reach is getting more frequent rather than less, that is the loop, and the loop is the finding.
And one more line from the policy statement, which I did not expect to see in print:
> "Societal framing around digital media use can incite caregiver guilt."
Good. Guilt is a terrible instrument. It makes parents defensive about the entire category, which means they defend the flight and the DMV line and the sick day with the same energy they defend the 6pm meltdown, when only one of those is the thing the evidence is actually pointing at.
None of this is a judgment about your particular child, and it should not be read as one. Your pediatrician is the only person with standing to say anything about your individual kid, and that has not changed.
Why we built Prodigy around the moment, not the metric
The reason I keep coming back to this finding is that it is a behavior, not a number. You cannot see it in a weekly screen time report. It shows up as a four second decision in a grocery aisle, and the only useful help is help that arrives before the decision, not in a summary afterward.
That is what Prodigy is for. It learns your child's actual patterns, and when a hard moment is predictable, it gives you something specific to try instead. The rep goes back where it belongs.