How I use AI to bring my kid's art to life – and why it's a fun learning opportunity

Follow ZDNET: Add us as a preferred source on Google.

---

ZDNET’s key takeaways

• AI can enhance kids’ art without replacing their creativity.
• Starting with original drawings keeps the child as the artist.
• Simple prompts help kids learn ethical, hands-on AI use.

---

I write a lot about AI tools for work, and I’ve been trying to figure out how to introduce my 4-year-old daughter to them without turning AI into a magic “make stuff for me” button.

I recently saw a CBS Sunday video of actor Ethan Hawke saying he’s “so bored” by AI, and that one thing he loves about theater is that AI can’t do it. 

“I feel I couldn’t be less interested in computers and fake things,” he said. “I like people. I like the way they smell. I like the way they talk, and I like the way they think. I like to think of AI as a plagiarizing mechanism. You know, that’s all it is.”

He also said, “I know it’s going to change the world, and it’s screwing everybody up, and I’m not in denial about that. But I’m in open rebellion.” As a writer first and foremost, I couldn’t agree with him more. But as a tech journalist, I also know for a fact that AI is already changing the world, and I want my daughter to be prepared for that reality.

Also: Free Photoshop in ChatGPT: How to edit photos with AI, no experience necessary

At the same time, I want her to stay creative, messy, imaginative, and hands-on. So, I’ve been showing her how to use ChatGPT, Gemini, and OpenAI’s Sora to augment her own drawings instead of generating art from scratch.

As critics have pointed out, AI can only generate what it’s been trained on, which is other artists’ work. By starting with ideas she actually put on paper, it turns AI into a fun learning tool while reinforcing that she’s still the artist.

(Disclosure: Ziff Davis, ZDNET’s parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

1. Use ChatGPT to make 3D illustrations

I like to give my daughter paints, markers, crayons, and canvases, and let her have fun nearly every day. Sometimes she creates quick illustrations, and other times I’m genuinely impressed by the level of detail she adds.

Either way, I want to encourage her to make whatever she wants. So when I use ChatGPT on her artwork, the goal isn’t to replace her style but to bring to life what’s already there. For example, she recently painted a simple green and purple character with orange eyes. I took a photo of it and uploaded it to ChatGPT to see what it could do with her art.

Since I wanted to keep her composition, colors, and shapes but add texture, depth, and detail, I used the following prompt:

Bring my daughter’s drawing to life as a more realistic, living character. Keep the feel of her original artwork, including its colors and shapes, but add texture, depth, and soft 3D detail so it looks alive and tangible while still clearly being her art.

The final result still looks unmistakably like her drawing, just more dimensional and realistic. When I showed it to her, she was amazed and immediately wanted to do more. She couldn’t wait to paint again, so we’d have something else to upload.

How to use ChatGPT on your child’s art:

1. Open ChatGPT on desktop or the mobile apps and start a new conversation.
2. Optional: Enable temporary chat if you don’t want your art to train future AI models.
3. Select the Create Image tool in the prompt field.
4. Upload a photo of your child’s artwork.
5. Use the prompt above or something similar.
6. Adjust the result with follow-up prompts, if needed.

Also: Is that an AI image? 6 telltale signs it’s a fake – and my favorite free detectors

This is also where your kid can get involved. I read my original prompt out loud to my daughter, but after the initial generation, I let her decide what else to add. She wanted to see her character with red eyes, sparkles, and a smile, so we experimented from there.

2. Use Gemini to reimagine characters

On another painting, I used Gemini Nano Banana Pro, which does a great job of retaining likeness if you upload a photo of an actual person and ask the AI to manipulate it while keeping their face and expressions intact. Now, you could use the same prompt above that we used for ChatGPT, but I’m trying to also show her how to play with prompts. So we went with:

Make my daughter’s drawing “come alive” as a more life-like, realistic version of her character. Keep the original color palette and composition so it is still clearly her art, while gently reimagining it.

Her original painting showed a red and green creature with what looked like arms and legs. Gemini leaned into that idea and turned it into a more defined character while preserving the same colors and composition. The first generation actually created a tree-like figure, complete with a trunk, leaves, and apples, but it had eyes, and the branches became the arms and legs.

How to use Gemini on your child’s art:

1. Open Gemini on desktop or the mobile apps and start a new conversation.
2. Optional: Enable temporary chat if you don’t want your art to train future AI models.
3. Select the Create Image tool in the prompt field.
4. Upload a photo of your child’s artwork.
5. Use the prompt above or something similar.
6. Adjust the result with follow-up prompts, if needed.

Also: Inside the making of Gemini 3 – how Google’s slow and steady approach won the AI race (for now)

My daughter thought the first generation was hilarious but quickly made sure to tell me, “No, Mommy, it’s supposed to be a lizard!” So we asked Gemini to make it more lizard-like and had fun tweaking the prompt to get different results.

3. Use Sora to animate artwork

This is where things really got entertaining, and it’s the AI tool my daughter had the most fun trying. We took one of her paintings, which she said is a “blue guy eating a caterpillar.” Right. Anyway, we fed that into Sora. But instead of asking Sora to add 3D detail or reimagine it, I created a very simple prompt. And since Sora includes audio, I asked it to make a noise too. We tried:

Make my painting come alive and be realistic. The blue guy should eat the caterpillar and say, “Mmhmm.”

I didn’t mention my daughter in the prompt like I did in the other examples because, in my experience, Sora flags any mention of children as a potential content violation and won’t generate a video.

How to use Sora on your child’s art:

1. Open the Sora app for iOS or Android and tap the + Generate button at the bottom.
2. Select the Upload Photo button next to the prompt field.
3. Upload a photo of your child’s artwork.
4. Enter a prompt describing the movements and sounds you’d like to see, and mention a style like photorealistic or illustrated.
5. Adjust the result with follow-up prompts, if needed.
6. You do not need to publically share your video and can simply store it as a draft.

Also: Stop accidentally sharing AI videos – 6 ways to tell real from fake before it’s too late

Our finished result? You can check it out for yourself here. It’s a little unsettling, but my daughter got a huge kick out of it and has probably rewatched it 50 times. She now asks me to use Sora on nearly everything she creates and is getting better at giving specific prompt instructions, like making her characters say something or move a certain way.

Why this is different from ‘AI art’

A lot of people are understandably critical of AI image/video generators, especially when it comes to artists’ work being used without consent. I get that, and that’s why teaching my daughter about AI early, and how to use it ethically and thoughtfully, matters.

Here, the starting point is my daughter’s own original art. We aren’t asking AI to copy someone else or generate something out of thin air. She’s learning how AI can build on her creativity, not replace it. With this approach, she’s still the artist, and the AI is a tool, not the creator. That distinction is especially important to me when introducing her to this technology.

Also: Hate creating presentations? I tried Google’s new Nano Banana Pro-powered tool, and the results blew me away

The thing is, this isn’t just for kids. If you sketch, paint, doodle, or design characters, you can use the same workflow to find inspiration for your own artwork. Upload your illustrations and tell the AI to respect what you’ve made. Use prompts that emphasize retaining likeness, colors, shapes, composition, and even style, and be as specific as you need to be while experimenting.