As kids, many of us felt the sting of a bully’s insults or were targeted by gossip. But these days, kids are putting mean and hurtful words in texts and on social media where they can linger. Many are also taking or forwarding pictures of classmates without consent. And if that’s not scary enough for parents, artificial intelligence (AI) has now made it possible for people to create fake images of real people, which are called deepfakes. And deepfakes are a risk we need to be aware of because they could affect our kids. Do we know how to keep our kids safe from AI?
“A deepfake is a video, photo, or audio recording that seems real but has been manipulated with AI,” according to GAO.gov, the Government Accountability Office. And some middle school kids in Beverly Hills, California, according to a report by NBCNews, have already been caught creating and sharing AI-generated nude photos of classmates. Times have changed since we were in school. Here are 5 ways to keep our kids safe in a quickly changing tech world with AI-generated images and deepfakes.
1. Stay informed about what people are using AI to generate.
If we want to know how to keep kids safe from AI, we have to pay attention. The five eighth graders in California who created nude photos of classmates were expelled. But the creation of deepfakes and the humiliation it has caused kids who’ve been targeted has continued to happen in schools across the country. Student-generated deepfakes have been reported in New Jersey, Florida, Washington state, and California. “Twenty states have passed laws penalizing the dissemination of nonconsensual AI-generated pornographic material,” says Politico journalist Dana Nickel, but how school districts handle the discovery and reporting of the deepfakes to police varies by state. Los Angeles Times journalist John Healey says there are dozens of AI-powered apps “available online to ‘undress’ someone in a photo, simulating what a person would look like if they’d been nude when the shot was taken.” Creating these images can cause a great deal of pain and embarrassment to peers and have a profound and devastating effect on victims.
Bottom line: Talk with your kids in an age-appropriate way about AI-powered apps and their capabilities. Tie it in with respect and kindness for others.
2. Educate your kids about genAI with your values in mind.
“What do you guys know about generative AI?” I asked at dinner. “Or ChatGPT?” My two kids didn’t offer much, but they had heard of it. So, we shared what we knew. Because genAI technology is currently being developed and released at a rapid pace, we as parents have a unique opportunity to learn with our kids or even stay one step ahead. There have been so many advances in the world of genAI in the past 18 months that it can be hard to stay on top of it all. But at this moment, teaching kids about genAI, while at the same time emphasizing your family values, is crucial.
Bottom line: Generative AI can create new writing and images from scratch. Talk with your kids about whether passing off AI-generated work as their own is the right thing to do.
3. Establish clear tech rules before letting your child use a chatbot.
Hindsight is always 20/20. But we can stay ahead of the game and keep our kids safer with technology by being clear up front about rules. As New York Times journalist Jessica Grose argues, “Kids need to be better educated, starting in elementary school, about technology and consent” before incidents like the one in California happen. She continues saying that these days, it is “typical for kids to get their own cellphones at around 11 or 12, and many kids even younger than that have access to a family iPad with image creation and sharing capabilities.” If kids have the tech already in their hands, this makes having a conversation about consent and deepfakes even more important.
Bottom line: Kids should be clear on the rules about not taking or forwarding images without the other person’s permission.
4. Try out an AI image-creator together before handing over the tool.
AI has made it easy to generate new images within seconds. It’s a powerful tech tool and your kids might be excited to try it, but also not ready to handle it. My daughter and I recently asked an image creator for “a dragon drinking out of a pond.” We received some cool-looking dragons that could’ve starred in a picture book but also some snarling dragons with razor-sharp teeth. While you may think pictures of dragons are somewhat innocent, these AI tools can create images your kids aren’t ready for.
AI image creators are still learning and may also mess up. There have been instances of AI-generated images showing people with too many fingers or teeth. Recently, Gemini’s AI image creator created inaccurate historical figures, and the company paused this feature to make improvements.
Bottom line: Companies are still fine-tuning AI technology. Chatbots have been known to get things wrong with facts and the same thing goes for images. Keep this in mind, along with the age of your child, before you give her the tool.
5. Take steps to protect your kids from predators who use AI tech.
When you’re wondering how to keep kids safe from AI, rest assured that there are steps you can take. You might’ve heard about the deepfakes involving Taylor Swift, President Biden, and President Trump. They fooled a lot of people. But our kids are susceptible to the same exploitation if their pictures are online. The South Carolina Attorney General’s Office said that “[C]reators are taking identifiable photos of children and morphing them into pornographic images through the use of AI, deepfakes and other editing software.” Photos of your kids on Facebook or TikTok are susceptible. The good news is lawmakers are introducing bills to protect kids, but it still might be something to think about if your family has a public social media presence.
Bottom line: Now that powerful AI capabilities are here to stay, you may want to rethink what photos you post online. With that, double-check your security settings so only those you trust can view your pictures.
What advice would you give to somebody who wants to know how to keep kids safe from AI?