Baby Suite
← All posts
A baby monitor used to be a microphone and a speaker. Then it grew a camera. This summer it grew ambitions. Nanit, the best-known of the AI monitors, raised $50 million this summer to take its cot camera beyond sleep and into tracking "speech and language development, motor skills and more", with a stated plan to keep watching "into early adolescence" (The New York Times, 2 August 2026, summarised by Bruce Schneier and Boing Boing). The company says it has around a million daily users. In the Netherlands the Pro camera with its floor stand lists at €399 (Nanit NL), with a six-month trial of the Insights plan and a subscription after that; in the United States the kit runs to about $474 and the "smart" features sit behind a plan of roughly $100 a year, call it €90 (The Hustle). Marketplace called it a cottage industry of AI subscriptions moving into the nursery. The story was written in New York; the devices are on the shelf in Amsterdam, Milan and Paris, and the questions are the same.
None of that makes these devices bad. Some parents sleep better with one; some babies are born with reasons for closer watching. But a camera that scores your child is a different purchase from a camera that shows you your child, and the store page rarely explains the difference. This post does: what the devices collect, what the research says about whether they help, and six questions to ask before one goes over the cot.
Start with the obvious: video and audio of a bedroom, around the clock. Not only the baby: the sibling who wanders in, the grandparent who does the night feed, you at 3am in whatever you sleep in. Most cloud monitors send that feed to the company's servers so the app can show it on your phone anywhere, and so the company's software can analyse it.
Then the derived data, which is the part that changes over time: sleep timings and "quality" scores, breathing motion, position and face-covering alerts, cough detection, and now, in the next generation, developmental scores computed from how your child moves and babbles. Wearables add heart rate and blood-oxygen readings (Owlet's Dream Sock, €299.99 in the Benelux, about the same in dollars). One camera advertises "AI-scheduled lullabies" (CuboAi, $359, roughly €330). Every one of these is a time-stamped record of your child that did not exist before, stored by a company you have a subscription with.
The companies say the data is "secure and is never sold or used for marketing". Take them at their word today and the harder question still stands, the one Schneier asks: what happens to a decade of bedroom footage and development scores when the company is sold, runs out of money, or changes its terms? A promise is not a deletion.
The most useful study we know of looked at how these devices behave in real homes rather than in a lab. A Canadian research team analysed 11,262 online reviews of 48 infant sleep monitors (wearables, cameras and under-mattress sensors) and published the results in BMJ Paediatrics Open in 2025. Three findings matter for a buying decision:
On the medical question, the position of the American Academy of Pediatrics is clear and worth quoting rather than paraphrasing: its 2022 safe-sleep guidance says "do not use home cardiorespiratory monitors as a strategy to reduce the risk of SIDS", and that a family's decision to use monitors "should not be considered a substitute for following AAP safe sleep guidelines". Consumer heart-rate and oxygen monitors are not forbidden; the guidance simply notes there is no evidence they prevent sleep-related deaths, and that they can create a false sense of security. European guidance says the same in plainer words. The UK's Lullaby Trust: "there is no research showing they help prevent SIDS", and "a monitor can never replace a parent or carer's supervision of a baby". The Dutch Consumentenbond (updated July 2026), citing VeiligheidNL, goes further and advises against breathing and movement sensors for healthy babies altogether, because of false alarms and false reassurance; medical monitoring is for babies whose doctor prescribes it. If a device is being sold to you as safety, those are the sentences to hold it against. For anything about your own baby's health, your doctor, midwife or consultatiebureau is the person to ask, not a camera and not a blog post.
And one question for yourself rather than the manufacturer: what would a "development score" from a camera change? If the answer is "I'd ask at the consultatiebureau", or the health visitor, or the paediatrician, you already have that person, and in most of Europe the visits are free and on a schedule. If the answer is "I'd worry", the score is selling you the worry. The professionals who assess a child's development do it in person, with the whole child in front of them, and they do not charge €90 a year.
a monitor should make you calmer, not busier ✎
We build a baby tracker, not a monitor, so here is the honest version of where we stand on the questions above. Baby Suite has no camera, no microphone and no wearable; it only knows what you type or tap. We store no photos or video anywhere in the suite, by decision rather than by accident (why, in our FAQ). Our AI features do not score your baby or give a cause for anything: the Rhythm check shows how the last few hours compare with her own usual rhythm, from your own logs, and every AI-written line is labelled as such. When a subscription ends the Free plan simply continues: nothing is deleted, export keeps working, and deleting your account is a button inside the app that works on the spot. Data is stored in the EU. Where we're not a substitute: if you need medical-grade monitoring for a baby with a diagnosed condition, that is a conversation with your paediatrician, and a prescription, not an app store.