Family photos help us stay connected, but AI and new digital technologies are changing.
Sharing family photos has always been about staying connected. A new baby, the first day of school, a birthday, a family vacation or simply an ordinary moment that a grandparent would love to see can be shared almost instantly with relatives and friends, regardless of where they live. Social media made that extraordinarily easy, and for many families it became the default place to share their lives.
What has changed is not our desire to share those moments, but the digital environment surrounding them. Artificial intelligence, facial recognition, large-scale data collection and increasingly sophisticated image-generation technology have introduced considerations that barely existed when many of today’s social platforms first became part of everyday life. A photograph placed online is no longer necessarily seen only by the people we imagined when we posted it. Depending on where and how it is shared, it can potentially be copied, indexed, analyzed, redistributed or incorporated into datasets outside our control.
None of this means families should stop sharing photographs. The more useful question is whether every photograph we want to share needs to be posted. There is an important distinction between publishing a family photo through a social platform and privately sharing that same photograph with the people we actually intended to see it.
Sharing family photos safely means considering who can access them, what personal information they may reveal and how images could be stored, copied or reused. Private albums, restricted-access sharing and careful decisions about what to post can help families stay connected while protecting their privacy.
What a Family Photo Can Reveal
The term “sharenting” is commonly used to describe parents sharing information, photographs and videos of their children online. The motivation is usually understandable. Parents are proud of their children, and photographs allow grandparents, relatives and friends to participate in their lives. UNICEF’s guidance on sharenting acknowledges this benefit, noting that sharing can help family and friends remain connected to important moments in a child’s life.
The privacy concern begins when we consider how much information an apparently ordinary photograph can contain. A first-day-of-school picture, for example, might reveal a child’s face and approximate age, while a school uniform, sign or building identifies where the child attends school. A caption may provide a name, grade or teacher, and details visible in the background can sometimes reveal a neighbourhood or home address. A succession of seemingly harmless posts can reveal considerably more when viewed together.
Ireland’s Data Protection Commission illustrated this problem through its Pause Before You Post campaign. In one example, information parents had shared online could be pieced together to reveal details including their daughter’s name, age, date of birth, football club, training schedule and information about when her father collected her. The individual posts may not have appeared particularly sensitive when they were originally shared, but collectively they created a detailed picture of the child’s life.
Canadian privacy researchers have examined similar concerns. Research funded by the Office of the Privacy Commissioner of Canada has considered the implications of sharenting for children’s privacy and image rights, including unwanted redistribution of photographs, identity-related risks and the creation of a digital footprint that children themselves did not choose.
This is an important distinction because the concern is not that every family photograph creates an immediate danger. Rather, it is that photographs are pieces of personal information, and the context surrounding them can reveal much more than we initially intended.
How Artificial Intelligence Changes the Conversation
These questions have become more significant as artificial intelligence has developed.
Manisha at AI Family Network explores this particularly well in her article, Sharenting Risks in 2026: What Parents Need to Know About AI. Her argument is not simply that parents should stop sharing photographs. Instead, she considers how the technological environment around those photographs has changed and encourages families to think more carefully about where children’s images are being placed and who—or what—may ultimately have access to them.
There is documented evidence behind some of these concerns. Human Rights Watch investigated LAION-5B, a very large image-and-text dataset assembled from material available on the web and used in artificial-intelligence research and development. Its investigation identified photographs of children within the dataset, including images that could be connected with names or other identifying information.
Research from the OECD examining data-collection mechanisms used for AI training raises a related issue. Web scraping can collect publicly accessible personal information even when the person represented by that information was not responsible for posting it. A child’s photograph uploaded by a parent, relative, school, club or other third party can therefore create privacy implications for someone who never made the decision to publish the image.
It is important not to exaggerate what this means. Posting a photograph online does not automatically mean that it will be used to train an AI system. Different platforms have different policies and practices, and the accessibility of an image varies considerably depending on where and how it is shared. The broader issue is one of control: once personal information is made publicly accessible, it becomes much more difficult to know where it may eventually be copied, indexed, analyzed or reused.
That loss of control matters even more when the person in the photograph is a child who had little or no say in the original decision to publish it.
When a Photograph Becomes Something Else
One of the most striking demonstrations of how rapidly this technology has developed is Deutsche Telekom’s A Message from Ella, produced as part of its #ShareWithCare campaign.
The campaign uses a fictional nine-year-old girl named Ella. With her parents’ permission, an image of the child was used with artificial-intelligence technology to create an older, digitally generated version of her. The adult “Ella” then confronts her fictional parents with examples of what increasingly accessible technology could potentially enable when children’s images are made available online.
[EMBED OFFICIAL “A MESSAGE FROM ELLA” VIDEO HERE]
The film is deliberately dramatic, and it should be understood as an awareness campaign rather than evidence that every publicly shared photograph will be misused. What makes it effective is the question underneath the dramatization: when we publish an image of a child today, can we confidently predict what technology may be capable of doing with that image five, ten or twenty years from now?
That question has become increasingly relevant. In February 2026, the Office of the Privacy Commissioner of Canada joined international privacy regulators in issuing a statement concerning AI systems capable of generating realistic images and videos of identifiable individuals without their knowledge or consent. The regulators specifically identified potential harms involving children and other vulnerable people.
The technology will continue to change, which makes it difficult to base good privacy decisions solely on what is possible today. A more durable approach is to think carefully about how much personal information we make publicly accessible in the first place.
Children and Their Own Digital Identity
There is another side of this discussion that has very little to do with artificial intelligence: the child’s own right to develop a digital identity.
Very young children obviously cannot understand what it means for a photograph to become part of an online history that may persist for years. As children become older, however, they can increasingly participate in decisions about what is shared about them.
UNICEF recommends considering a child’s perspective and, where age-appropriate, involving children in decisions about sharing their images. These conversations can also provide an opportunity to teach something broader: photographs are personal information, and respecting someone’s image is part of respecting that person’s privacy.
A useful habit, therefore, is simply to ask before posting whenever a child is old enough to participate in the decision. When children are too young to answer, parents can still consider whether the future version of that child is likely to be comfortable with the image becoming part of a publicly accessible record.
Sharing Does Not Have to Mean Posting
One of the most useful ideas in Manisha’s AI Family Network article is her PHOTO Check, a framework designed to encourage parents to pause before sharing children’s images. Rather than reproducing the framework here, her original article is worth reading in full. It considers permission, identifying information, the intended audience and alternatives to public posting, including private albums, invite-only groups and encrypted sharing.
That last point deserves more attention because we often use the words posting and sharing as though they mean the same thing. They do not have to.
Posting generally begins with a platform: we open a social network, upload something to a profile or feed and then determine which portion of that platform’s audience is allowed to see it. Private sharing can begin from the opposite direction. Instead of asking where we should post a photograph, we first ask who the photograph is actually for.
Perhaps it is for grandparents and immediate family. Perhaps it is for several close friends. Perhaps it is for everyone who attended a wedding, family reunion or vacation, but nobody outside that group. Once the intended audience is defined, we can choose a method of sharing that reflects that relationship rather than automatically placing the photograph into a broader social environment.
That might mean sending a photograph directly to someone, maintaining a private family album, using an encrypted photo-sharing service or creating an invite-only group. None of these approaches can eliminate every possible risk. Anyone who is permitted to see a photograph may still be able to save it, copy it or photograph the screen. But restricting access from the beginning is fundamentally different from making the same photograph broadly accessible and then attempting to control what happens to it afterward.
Making More Deliberate Choices About Family Photos
Protecting privacy does not require families to become cybersecurity experts. In many cases, it simply requires developing a few better habits around what is shared, where it is shared and who can see it.
Before sharing a photograph, look beyond the people in the picture. A school logo, street sign, house number, vehicle licence plate, sports jersey, name tag or document sitting in the background may reveal information that was never intended to be part of the post. Captions deserve the same attention. Combining a photograph with a child’s full name, birthday, school, location or regular activity can turn an otherwise ordinary image into a much more detailed record.
The audience matters just as much. If a photograph is really intended for ten relatives, there is little reason for hundreds of social-media connections—or potentially people beyond that network—to have access to it. Privacy settings can help when social media is used, but those settings should be reviewed rather than assumed. Families should understand who can see what they share and, where possible, what the service itself does with uploaded information.
It is also worth becoming comfortable with the idea that not everything needs to be online. Some photographs can remain within the family. Others can be printed, placed in photo books, sent directly to relatives or shared through restricted-access services. Privacy is not an all-or-nothing choice between publishing everything and sharing nothing. There is a considerable amount of room between those two extremes.
Privacy Doesn’t Have to Mean Isolation
Discussions about children’s privacy can understandably become alarming. Artificial intelligence, deepfakes, facial recognition, data collection and identity theft are serious subjects, and focusing exclusively on the risks can make withdrawing from online sharing altogether seem like the only responsible option.
But that overlooks something equally important: family photographs have genuine social value. They allow grandparents to watch grandchildren grow up even when they live thousands of kilometres apart. They let extended families participate in birthdays, holidays and ordinary moments they would otherwise miss. They preserve memories and give people a way to remain present in each other’s lives despite distance.
The objective, therefore, does not need to be to stop sharing. A more practical objective is to share with intention.
Understand what is visible in the photograph. Think about the information accompanying it. Consider the privacy and wishes of the people pictured. Decide who actually needs access. Then choose a method of sharing appropriate for that audience.
Technology has made it extraordinarily easy to publish our lives. Perhaps the next step is becoming better at deciding which parts of our lives actually need to be public.
Frequently Asked Questions
Is it safe to share family photos online?
It can be, but no online sharing is completely risk-free. Limit who can see your photos and avoid sharing unnecessary personal information.
Can photos posted online be used to train AI?
Publicly accessible images can potentially be collected for AI datasets. Posting a photo does not automatically mean it will be used for AI training, but public images are harder to control.
What is sharenting?
Sharenting is when parents or caregivers share photos, videos or information about their children online. Over time, this can create a digital footprint for the child.
What information should I avoid showing in children’s photos?
Avoid revealing school names, addresses, street signs, licence plates, schedules, name tags and other details that could identify a child or their location.
Should I ask my child before posting their photo?
When they’re old enough to understand, yes. Asking gives children a say in their privacy and helps teach them about consent.
What’s the difference between posting and privately sharing photos?
Posting usually places a photo on a social platform or feed. Private sharing limits access to specific people through direct sharing, private albums or invite-only groups.
What’s the safest way to share family photos privately?
Share only with people you trust using restricted-access or invite-only services, and avoid including unnecessary identifying information.
A Note From Tapestry
These ideas are closely related to why my brother and I co-created Tapestry Show. We wanted family and friends to be able to share photographs with the people they choose without requiring those memories to become part of a public social-media presence.
Tapestry uses private, invite-only Shows and was designed without public profiles, advertising or tracking. It is one option for families who want to keep sharing photographs while being more deliberate about who receives them.
Regardless of which service you choose, however, the more important habit is the one that applies before any photograph is uploaded: pause for a moment and consider who you are sharing it with, what else the photograph reveals, and whether this is the right place to share it.
