How much screen time is too much? A by-age reality check

What the guidance actually says

• Under 18 months: video calls with family only.
• 18–24 months: high-quality content, watched together, briefly.
• Ages 2–5: about one hour a day of high-quality programming on weekdays, co-viewed where possible.
• Ages 6–12: no fixed official number. Roughly 1–2 hours recreational on school days is where most paediatric advice lands, with sleep, activity and schoolwork protected first.
• Teens: totals stop being the useful measure. Judge by what the screen is displacing.

Note what the guidance does not do: it does not distinguish an hour of a chatbot from an hour of cartoons. That distinction is now yours to make, and it's the subject of our screen time rules for kids in the AI age.

The five signs that matter more than the number

• Sleep. Later bedtimes, harder wake-ups, or a device in the bedroom overnight. This is the first domino.
• Mood on exit. Ordinary reluctance is fine; genuine distress or rage when a screen ends is a signal.
• School. Homework that arrives finished but unexplainable, or work that stops getting started at all.
• Friendships. Online contact replacing in-person contact, rather than adding to it.
• Interest collapse. Things they used to love — sport, drawing, an instrument — quietly stopping.

If none of those five are moving, your family is probably fine even if the total looks high. If two or more are moving, the total is too high regardless of what it is.

Imagine the weekly total is lower than last month, but bedtime keeps slipping. Look at the timing as well as the amount. Ask whether homework, messages, or entertainment is pushing the evening later and what could move earlier. A total doesn't show when sleep was crowded out. At the same time, tiredness can have other causes, so don't assume a screen report explains everything.

Suppose the teenager stopped attending a hobby they used to enjoy. Ask what changed before deciding the phone caused it. They might have a scheduling conflict, a disagreement with someone there, or a different interest. The screen pattern belongs in the conversation, but it isn't the only evidence. You can notice a change without turning it into a label about dependence.

For a weekend example, imagine the child watches a long film with cousins and then spends the afternoon outside. Compare that with shorter sessions that repeatedly interrupt meals or homework. The contexts are different even if the timer favors one day. Ask what the screen activity displaced and whether the family responsibilities were met, rather than treating every unusual day as proof the plan failed.

Suppose stopping consistently produces distress. Note which activity, time of day, and transition are involved. An abrupt stop during a game may need a different plan from late-night open-ended chatting. The observation gives you something specific to adjust or discuss with a trusted professional if needed. It doesn't establish the cause from the reaction alone.

The AI-specific version of "too much"

There's a newer signal worth adding: work that is finished fast and can't be explained out loud. That's not a screen-time problem, it's an outsourcing problem, and it responds to a different fix — the signs are in is AI doing your kid's homework.

When you're unsure about a specific assignment, paste it into the AI homework checker for a plain-language read on whether it shows real thinking. If they're stuck rather than shortcutting, explain my kid's homework gets them moving again in a few minutes instead of an evening.

Imagine a child asks for help before attempting even a familiar practice problem. Ask what they would try first with the screen closed. They may be confused, afraid of mistakes, or accustomed to immediate reassurance. Those possibilities need different support. The useful next step is one independent attempt at an appropriate level, not a lecture about never needing help.

Suppose an eligible teen spends a long time discussing an essay but still has no claim of their own. Ask what the assignment requires and have them choose one point from the reading offline. Length of conversation isn't the same as progress. A short pause to identify the missing decision may help you see whether the tool is supporting the task or delaying it.

For an adult-mediated learning activity, imagine you use your eligible account to understand a general concept, verify it, and then explain it offline to your child. Finish with a separate example they can attempt themselves. That gives you something to observe beyond time spent with the explanation. Keep account eligibility and teacher rules in view; a useful goal doesn't override either.

What to do with the answer

If you've decided it's too much, don't start with a number — start with the defaults. How to reduce screen time without a nightly fight covers the changes that hold, and our plans give each child in the house their own profile if you're managing different rules for different ages.

Suppose you decide one routine needs changing. Choose a specific experiment, such as moving devices out of bedrooms at the agreed time, and decide when you will review it. Record what actually changes and what doesn't. Don't promise that a single rule will fix mood, sleep, or schoolwork. The review is a way to learn about this family's routine.

Imagine siblings respond differently to the same plan. Ask what each activity involves and what support each child needs to stop or begin something else. Keep the reasons transparent so flexibility doesn't feel arbitrary. A changed schedule can be fair when its purpose is clear and the basic boundaries remain understandable.

Suppose worries about sleep, distress, or relationships persist beyond the screen plan. Seek appropriate human support rather than trying to infer a diagnosis from usage minutes. Bring concrete observations if helpful: when the difficulty occurs, what activities are involved, and what you've tried. A report of the pattern is more useful than a claim that a number proves the problem.