What AI Does With Uploaded Images & Data (My Research Findings for Zazzle Creators)

[Susang6]

A lot of Zazzle creators have been asking me what different AI tools actually do with uploaded images, mockup images especially when it comes to AI training data and digital copyright, so I wanted to put everything I’ve learned in one place. I’m not an authority or anything like that I just read a lot, and because I’m a full‑time caregiver, I tend to do my research late at night when the house is quiet. It helps me stay focused and calm, and when someone asks me a question about AI, copyright, or privacy settings, I like to go find the answer.

One of the biggest questions I get is whether uploading an image into an AI tool means the AI takes it for training or affects your copyright ownership. With Microsoft AI tools, uploading an image does not transfer your copyright. Microsoft confirmed directly to me that I retain the copyright on my mockups and any uploaded images. Their privacy settings and documentation explain exactly how uploaded data is handled, and nothing is automatically added to AI training data. If anyone wants the official source, Microsoft publishes everything here: https://privacy.microsoft.com/en-us/privacystatement.

Since many Zazzle creators also use Google Gemini, I checked that too. I asked Gemini directly whether I keep my copyright when I upload my own images, and they told me yes  especially when the image is something I created myself. Google’s documentation supports that, stating that they don’t claim copyright over what Gemini generates and that users keep ownership of outputs created with their prompts. Their privacy settings explain that uploaded images are stored temporarily for processing and may be used to improve AI models only if you opt in. You can also manage or delete your data in your Google Account activity. When I was invited to be a beta tester for Google’s AI several years ago, they required an explicit opt‑in for data use, and I chose not to opt in after reading the fine print. And just yesterday, Gemini didn’t follow my instructions and the mockup came out terrible, and a human actually reached out afterward to see what went wrong, which I appreciated because it shows there are real people involved in quality control. Google’s official privacy hub is here: https://support.google.com/gemini/answer/13594961.

Since Canva is another tool a lot of creators use for mockups and design work, I looked into Canva’s privacy settings as well. Canva is very clear about how they handle uploaded images and AI training data. Your content on Canva is not used to train their AI unless you turn that setting on yourself, and that permission is off by default. You can check all of this in your Canva privacy settings  they let you control whether your content can be used to improve AI, whether your general usage helps them build features, and whether your design activity is shared with collaborators. Your designs are private unless you choose to share them, and Canva employees don’t automatically see your work unless you specifically share something with support. Canva’s official documentation is here: https://www.canva.com/help/manage-privacy-settings/ and their Trust Center privacy overview is here: https://www.canva.com/trust/privacy/.

I’m sharing all of this because I like to understand how things work, and when someone asks me a question about AI, copyright, or uploaded images, I want to give them the most accurate information I can find. Copyright still belongs to the creator, and Microsoft, Google Gemini, and Canva all publish exactly how they handle uploaded data. If anyone has questions or wants me to break anything down, I’m happy to help.

If you have any questions, just ask…or send me a message at
Susan’s Nature & Seasonal Studio

[Deb]

Hi.. reading your great articles on ai, marketing, mockups etc .. wow the research you have done , but wondering why some other programs are not included .. ie: Midjourney, DALL-E 3 (via ChatGPT/Copilot), Adobe Firefly, and Stable Diffusion? From what I can tell only canva , microsoft and gemini are used? Which of these are folks using for the mockups? Is there a reason these other platforms are not included in the mix?  Thanks .. and thanks for your great research !

[Susang6]

For this post, I focused on the AI tools creators had specifically asked me about. Those were Microsoft, Google Gemini, and Canva, so that’s where I put my research time. I know there are many other platforms like Midjourney, Firefly, and Stable Diffusion, but I don’t have the time to research those right now. Maybe another creator who enjoys that kind of research to see how things work like colescreations will take that on. My goal here was simply to answer the questions I’d already been asked, and I’m glad the information was helpful.