Your child’s latest birthday photo, that cute dance video. Share them on social media and they add to your family’s digital footprint. And depending on the settings you’ve chosen, AI could use them.
Dr. Siwei Lyu, a digital forensics expert and father of two, is a Distinguished professor of computer science at SUNY who studies how AI learns from public data. The danger of sharing anything online is that it becomes fair game for AI deepfakes: fake photos, videos or voices built from real images or recordings. Deepfakes can impersonate anyone, including kids and teens.
“A child’s data might be misused to recreate an image or video of them in situations that aren’t real,” Lyu explains. “That content can stay online and affect your child for years.”
Quitting social media isn’t necessarily the answer. Instead, Lyu tells his kids to share thoughtfully: post less, choose private settings and pause to consider what you’re sharing before uploading anything personal. Here, Lyu explains how your digital footprint can be manipulated into AI deepfakes.
How it happens: From real photos to AI deepfakes
Step one: The photo goes public. When your child uploads a photo or video, it’s part of the web. Even if they think it’s private, someone could take a screenshot, upload it and it could go online.
Step two: Manipulation happens. A person has to choose the image and upload it into an AI deepfake tool. That’s when it crosses from simple image capture to digital manipulation.
Step three: It’s shared. Once the AI deepfake is created, it can spread fast—shared on social platforms, sent in private chats or even posted to websites that traffic in AI-generated or altered images. That’s when it can become harmful or invasive.