For 14 months, writer and mom Jenny Rowe-Patel shared cheerful snapshots of her baby daughter online, until new evidence about AI-generated child abuse pushed her to rethink where those images belonged.
“But normal and safe aren’t necessarily the same thing,” Rowe-Patel wrote in a Marie Claire UK essay published Sept. 8.
Her old approach will sound familiar to plenty of parents. She announced her daughter’s birth on Instagram, posted vacation and first-birthday photos and gave the child’s nursery permission to use her image, all without much concern.
The unsettling turn came when Rowe-Patel read the Internet Watch Foundation’s latest data. The child-safety organization identified 8,029 AI-generated images and videos depicting realistic child sexual abuse in 2025, including 3,443 videos, up from just 13 the year before.
Nearly two-thirds of those videos fell into Category A, the most severe classification under British law. The same IWF report says freely available tools can produce convincing deepfakes of a specific child from as few as 20 photos in about 15 minutes.
That does not mean every family photo will be stolen. It does mean the cost of copying and corrupting an ordinary image has dropped sharply, while parents may never know that a stranger saved it.
Related: A Playground With Pedophiles: Tim Tebow Gets Real About Online Predators
“An account can look like one thing, but be something — or someone — completely different,” Internet Watch Foundation Head of Marketing Tom Dyson told Rowe-Patel.
Dyson, who does not post photos of his 7-year-old son on social media, warned that private accounts still require trust. One hacked account, unknown follower or screenshot can move an image far beyond the audience a parent intended.
In July, the IWF and the United Kingdom’s National Crime Agency issued new guidance for parents and caregivers. They recommend reviewing privacy settings, limiting photos to a small “close friends” group, checking old posts and reconsidering image-consent forms signed for schools, clubs or sports teams.
Parents should look beyond a child’s face, too. A school uniform, house number, street sign or familiar landmark can reveal more than intended, and covering a face with an emoji does not erase those clues.
The danger reaches American families as well. The National Center for Missing & Exploited Children categorized more than 158,000 images and videos submitted through CyberTipline reports as AI-generated between January 2023 and December 2025, and it warns that these tools can fuel exploitation, bullying, fake accounts and sextortion.
Families who discover an explicit image made or taken when the person depicted was under 18 can turn to NCMEC’s free Take It Down service. The tool creates a digital fingerprint on the user’s device so participating platforms can detect and remove copies without receiving the image itself.
Rowe-Patel has now tightened her privacy settings, cut her follower list in half, restricted family photos to close friends and plans to withdraw the nursery consent she once gave. She also wants relatives to understand the family’s new boundaries.
“Does this really need to be on social media?” she now asks before posting, often deciding that a private family message will do the job better.
Parents do not need to panic or master every new AI tool. They can start with faithful stewardship: protect a child’s privacy, ask permission as children grow, keep online-safety conversations open and choose a smaller circle when sharing a joyful moment.
Read Next: This Chat App Your Child Probably Uses Exposes Them to Predators
Questions or comments? Please write to us here.
