Photo Upload on HeyOtto: Why Image Safety Matters for Kids
By Natalie Gibson, Founder and CEOLast updated:
AI image safety for kids: what parents should know when a child can share photos with AI, and how HeyOtto blocks harmful images and keeps parents in control.

Key Takeaways
- Photo upload on HeyOtto starts turned off. Only a parent can turn it on, and only for children on HeyOtto Pro.
- Every uploaded image is screened, including for CSAM, before Otto can respond.
- Flagged images are blocked. Parents are notified, and the safety team reviews them. When required by law, HeyOtto reports to NCMEC.
- Photos are encrypted in transit and at rest, kept for up to 30 days, and are not used to train AI.
- Screening is not a guarantee that every harmful image will be detected.
Children don't always have the words to explain what they're thinking.
Sometimes they have a picture. A homework question they want help understanding. Something they found outside. An object they're curious about. Or something they simply want to ask Otto about.
That's one of the reasons photo upload is now available on HeyOtto Pro.
But adding image uploads to an AI platform built for children isn't as simple as adding an upload button. We had to think carefully about what it means to let children bring images into an AI conversation, and what responsibilities come with that.
Because when you're building AI for kids, safety can't stop at the text box.
Children experience the world visually
A growing number of AI tools let users upload photos and ask questions about them. For adults, this might mean uploading a document, analyzing a chart, or asking AI to explain something in a photo.
For children, the possibilities are different:
- Homework help: A student might want help understanding a diagram in a science assignment. Otto asks guiding questions instead of handing over the answer.
- Curiosity: A curious eight-year-old might want to know what kind of insect they found in the backyard.
- Context for a question: A teenager might want to understand something they saw, or share an image to explain what they mean.
- Creativity: A child might want to edit a photo or turn a drawing into a story.
Images can make AI more useful, more accessible, and easier to use. But they also bring a different set of safety considerations.
Why image safety matters when children use AI
Text-based safety systems are not enough when users can upload images.
Images can contain sensitive personal information, inappropriate material, or content that should never be part of a child's AI experience. One particularly serious concern is child sexual abuse material (CSAM).
Platforms need to consider how this content is identified and handled, and what happens when an image contains something inappropriate, harmful, or unsuitable for a young user.
These aren't problems that can be solved by telling children what not to upload. The platform itself has a responsibility.
How HeyOtto approaches photo upload safety
As we prepared to introduce photo upload, we wanted safety to be part of the feature from the beginning.
Every image is screened before Otto ever sees it. This is part of our larger approach to building AI specifically for children and teenagers.
We think about safety across the entire experience, including:
- What children type. Conversations need safeguards designed around the age and needs of the child.
- What children upload. Visual content introduces additional risks and needs its own screening.
- How AI responds. The response matters just as much as the content that starts the interaction.
- How parents stay involved. Parents need meaningful visibility and tools to help guide their children's AI experiences.
No safety system is perfect, and screening should never be presented as a guarantee that every harmful image will be detected. But we believe these safeguards are an important part of responsible image-enabled AI. You can read more about our approach on our safety page.
Parents decide who can upload photos
Every family is different. The parent of a 16-year-old and the parent of an 8-year-old may make different calls, and both are right. So photo upload starts turned off for every child, and only a parent can turn it on.
To turn it on:
- Make sure your family is on HeyOtto Pro.
- Open your parent dashboard.
- Choose your child's profile.
- Turn on Photo upload.
Repeat for each child who should have access. You can turn it off anytime.
If an image is flagged
- The image is blocked. Otto doesn't respond to it.
- You're notified. Parents get an alert when an image is flagged.
- Our safety team reviews it. When required by law, we report it to the National Center for Missing & Exploited Children (NCMEC).
Where photos are stored, and for how long
- Up to 30 days. Photos are stored securely and then deleted.
- Never used to train AI. Your child's photos are not used to train any AI model.
- Encrypted. Photos are encrypted in transit and at rest.
- One exception. If a photo is part of a safety review, we may keep it longer, as needed to protect children and meet our legal obligations.
For the full details, see our privacy policy and Trust Center.
We're building more than an AI chatbot
At HeyOtto, we've always believed that children deserve AI designed for them from the ground up. Not an adult AI product with a few filters added afterward.
That means thinking about age-appropriate experiences, privacy, parental involvement, and safety at the infrastructure level.
Photo upload is a natural next step in making AI more useful for kids and teens. It gives them another way to ask questions, learn, and explore. But every new capability brings new responsibilities.
We don't believe innovation and child safety should be competing priorities. They should be developed together.
A more thoughtful way for kids to explore AI
We're excited to give children more ways to use Otto and bring their curiosity into the conversation. And we're just as committed to continuing to build the safety systems that support those experiences.
Because the future of AI for children shouldn't just be about what technology can do. It should be about what we're willing to do to make it safer for the next generation.
HeyOtto. AI for curious minds. Safety built in.
Key Terms & Definitions
- CSAM
- Child sexual abuse material. HeyOtto screens uploaded images for CSAM before Otto can respond to them.
- Photo upload
- A HeyOtto Pro feature that lets a child send an image into a conversation with Otto. It stays off until a parent enables it for that child.
- NCMEC
- The National Center for Missing & Exploited Children. When required by law, HeyOtto reports flagged images to NCMEC.
Sources & Citations
The CyberTipline is NCMEC's centralized reporting system for online child sexual exploitation, including reports from electronic service providers.
National Center for Missing & Exploited Children
FAQ
Frequently Asked Questions
Common questions about this topic, answered.
No. Only a parent can turn it on, from the parent dashboard.
Yes. Each child has their own setting, so you choose who has access.
No. Never.
People trying to get around safety systems read those details too. What we can promise is the outcome: every photo is screened, flagged images are blocked, and serious violations are reported.
Related Articles
Ready to Give Your Child a Safe AI Experience?
Try HeyOtto today and see the difference parental peace of mind makes.






