Skip to content
Try Raven →
All posts
PrivacyBy the Raven team8 min read

Sharing Children's Photos Online Safely

A crest on a jumper, a number on a bin, the same swings every Saturday. What a family photo really gives away, and the small habits that fix most of it.

Short answer

Sharing children's photos online is usually fine, and the risk sits in repetition rather than any single picture. School badges, house numbers, street furniture and the same park each week build a routine. Strip location data, keep identifying detail out of frame, vary the timing, and prefer closed groups.

Abstract dark composition: a row of small identical squares repeating along a faint grid, one square softly blurred out, on near-black.

The first day of term photograph is the most informative picture most families ever publish. A child stands on the doorstep in a new jumper, the crest is legible, the door number is over their shoulder, and the caption names the year group. Nobody has to open the file or read a hidden tag to learn the school, the approximate age and the street. All of it is simply in the frame, in daylight, at the front of the picture.

That is worth knowing, and it is not worth panicking about. Almost nothing bad happens to almost any family, and the useful response is a handful of small habits rather than a decision to stop photographing your children. What follows is the calm version: which details actually do the work, why repetition matters more than metadata, and the five or six changes that remove most of the exposure without removing the point of sharing at all.

What does a photo of a child actually give away?

Usually four things at once: an institution, from a uniform crest or a book bag; a home, from a door number or a distinctive gate; a neighbourhood, from bins, bollards and shopfronts; and an age, from the child. One frame can carry all four with no metadata involved.

Start with the uniform, because it is the single densest clue in the picture. A school uniform crest is deliberately unique — that is the whole purpose of it — and it is drawn to be read at a distance, in bold shapes and two or three flat colours. Those are precisely the qualities that survive heavy compression, so a badge only 60 px wide on a phone screen stays identifiable to anyone who knows the district. Naming a school narrows a family to a catchment area, often a couple of square kilometres, and implies an age band without anyone stating it. Book bags, PE kits, water bottles and coat labels all carry the same crest, and they appear in far more photographs than the first-day shot.

School signage does the same job more bluntly. The drop-off photo taken by the gate, the sports day picture with a banner in the background, the nativity shot with a display board behind the stage — each of them names the institution outright. So does the reflective strip on a book bag from a particular supplier, though that one is a specialist's clue rather than an obvious one.

Then there is the ordinary street. A wheelie bin with a house number painted on the lid is a common and complete answer. So is a numbered door, a recycling box in a council-specific colour, a distinctive gate, or a view along the road with two shopfronts in it. The general version of this — how kerbs, bollards, bins and post boxes quietly name a place — is set out in how street furniture gives away a city, and children's photographs are unusually full of it because they are taken outside the front door.

Why is metadata the smaller problem?

Because it is already handled in most cases and easy to handle in the rest. Large platforms strip the Exif block on upload, and switching off camera location tagging closes the gap for everything else. The visible layer of a photograph cannot be stripped, which is why it matters more.

Phone cameras write coordinates into the file using Exif, the tag block that has ridden inside ordinary JPEGs since 1995. It is genuinely useful and genuinely leaky, and it is worth turning off or clearing before you send anything outside the big platforms — messaging apps that forward originals, email attachments and shared folders often pass the file through untouched. The mechanics of removing it take about two minutes and are covered in how to remove location data from photos.

Once that is done, though, you have solved the smaller half. Nothing strips a crest, a door number or a recognisable park bench, because those are the picture rather than an attachment to it. A photograph with every tag removed can still be placed to a street by anyone patient, and increasingly by software. That asymmetry is the reason this article spends more time on framing than on settings.

Where does the real risk sit: one photo or a pattern?

In the pattern, almost always. A single picture is a moment and reveals a moment. Fifty pictures taken at the same swings, the same gate and the same Saturday morning slot describe a schedule, and a schedule is the thing worth not publishing.

This is the part that people underestimate, because no individual post looks like the problem. The park photograph is charming. The second one, three weeks later, is also charming. By the twentieth, an ordinary reader who was not even trying has learned that your family is at a specific playground most Saturday mornings, that the walk home passes a particular parade of shops, and that the school gate sits roughly 400 metres from that parade. Nobody deduced anything clever. They read a feed in order.

Assembling a picture from scattered public fragments is a recognised discipline with a formal name, open-source intelligence, and the uncomfortable part is how little skill the family version requires. It also explains why the fix cannot live in any single post. You cannot crop a pattern. You can only vary it, thin it out, or move it somewhere with a smaller audience.

The same logic applies to holidays, where the risk flips from routine to absence — a live feed from a beach also announces that a house is empty. That side of it is covered separately in traveling without oversharing your location.

What habits help when sharing children's photos online?

Six small ones: strip or disable location data, keep crests and door numbers out of frame, delay posts by a day or two, vary the places you photograph, use a closed group for anything detailed, and ask other parents before posting their children.

  1. Turn off camera geotagging once, rather than remembering to strip files later. It is a single settings screen and it covers every photo you take afterwards.
  2. Watch the crest. Turn the jumper, crop the badge, or take the first-day photo indoors against a plain wall. The picture is the same; the school is not in it.
  3. Check for a number. Door plates, painted bin lids, wall plaques and estate signs are the details that turn a neighbourhood into an address.
  4. Delay by a day or two. A sports day photo posted on Wednesday is the same photo. It is simply no longer a live position for a child.
  5. Vary the backdrop. If three quarters of your photographs come from one playground, the playground is now part of your public record. Rotate, or crop tighter.
  6. Use a closed group for the detailed ones. Grandparents want the whole set; the open internet does not need it. Narrowing the audience beats any amount of careful cropping.

None of these asks you to post less, which matters, because advice that demands abstinence is advice nobody follows past February. They are frictions of a few seconds each, and they compound. A feed with no crests, no numbers, no live timing and a varied backdrop is dramatically less useful to a stranger while looking, to everyone who actually cares about your family, entirely unchanged.

Should you ask before posting somebody else's child?

Yes, and asking once is usually enough to settle it permanently. Families set the line in very different places, and a party photograph with four children in it makes a publishing decision on behalf of three other households that were never consulted.

This is the etiquette question that causes the most friction and the least harm, and it is easily solved. Some parents post everything; others have decided their child's face will not be public until the child can weigh in. Both positions are reasonable and neither is obvious from the outside. A single message in a class group — happy to post the party photos, tell me if you would rather yours were not in them — resolves it for a year.

There is a legal shadow behind the courtesy, too. Article 8 of the GDPR sets the age at which a child can consent to online services at 16 by default, and lets member states lower it to no younger than 13 years. That rule governs services rather than parents, but it reflects a settled instinct: a child's ability to agree to their own visibility arrives late, and the pictures arrive early.

The station noticeboard test

When a decision feels borderline, one question does most of the work. Would I pin this photograph to the noticeboard at my local station, where anyone waiting could read it? Not a hypothetical global audience — the actual people who pass through your area every day, which is a far more realistic description of who eventually sees a public post than a faceless crowd of strangers on the other side of the world.

The test works because it is concrete. Most bath-time photographs fail it immediately. A birthday cake photograph passes without difficulty. A first-day-of-term picture passes once the crest is out of frame and fails while it is in. It also handles the pattern problem, since the honest follow-up is whether you would pin the same photograph up every Saturday for a year.

What about the photos already online?

Change the habit rather than attempting a purge. Remove only the handful showing a door number, a school gate, or a place your family still visits weekly. Older pictures from schools and houses you have left behind are close to harmless and rarely worth the afternoon.

Retrospective clean-ups fail for the same reason crash diets do. The task is enormous, the payoff is invisible, and the abandoned attempt takes the good intention with it. Ten minutes spent on the three or four genuinely revealing posts is worth more than a lost weekend spent on four hundred harmless ones, and the habit change is what actually alters the next two years.

It is also worth seeing what a photograph gives up before you decide, which is a use for the same visual geolocation tools discussed across the rest of this cluster. Uploading one of your own pictures to Raven and watching it read the architecture, road markings and plants is a quick and slightly humbling calibration exercise. The image is processed in memory and discarded, and the guess is entertainment-grade and often wrong — but if a toy gets within a region from the background alone, a determined person will do at least as well. The wider account of what a photograph exposes sits in AI and photo privacy.

Upload a photo you were about to post and see how much the background alone gives away.

Upload a photo →

The realistic goal is not a family that never appears online. It is a family whose photographs are a little less useful to a stranger than they were last year: no crest, no house number, no live timing, no single playground repeated fifty times. That is an afternoon of attention and then a few seconds per post, and it is the whole of the practical advice.

Frequently asked questions

Is it dangerous to post photos of my children at all?
For the overwhelming majority of families, no. The realistic concerns are mundane: an identifiable school, a visible home, and a routine that a feed makes legible over months. Each has a small, practical fix that does not involve posting nothing.
Which single detail should I keep out of the frame?
A uniform crest, if you have to choose one. It names an institution, implies an age band and narrows the catchment to a few streets, all at once. A house number is a close second because it converts a neighbourhood into an address.
Do closed groups and private accounts actually help?
They help a great deal, though not perfectly. Narrowing the audience removes the passing stranger entirely, which is most of the problem. Assume anything shared can be screenshotted by anyone in the group and post accordingly.
What if another parent posts my child without asking?
Ask them directly and privately, and expect it to be an oversight rather than a stance. Most people take a photo down within the hour. Saying it once, early, also spares you the same conversation after every party for the next five years.
Should I worry about photos the school itself publishes?
Less than about your own, because a school newsletter is already tied to the school and adds nothing a reader did not have. Most schools hold a consent form on file. If you would rather your child were left out of the website gallery, say so in writing at the start of term.

Sources

  1. School uniformWikipediaCrested uniforms are old and highly local: Christ's Hospital, founded in 1552, still dresses pupils in a recognisable Tudor coat, and modern crests remain school-specific by design.
  2. Art. 8 GDPR — Conditions applicable to child's consentgdpr-info.euSets the default age of consent for online services at 16, which member states may lower to no younger than 13.
  3. Open-source intelligenceWikipediaThe practice of assembling a conclusion from scattered public fragments — the shape a family feed takes after a couple of years.
  4. ExifWikipediaThe metadata standard published in 1995 that carries camera model, timestamp and GPS coordinates inside an ordinary photo file.

Reminder

Raven is built for entertainment and curiosity. Its guesses are AI estimates that can be wrong, and it must never be used to track or identify real people. Uploaded photos are processed in memory and immediately discarded — never stored.

Get Geospy AI for iPhoneDownload free