How We Keep Stories Safe for Your Child

Whenever we mention that Wisp uses AI to generate stories, we get the same question almost immediately: "But how do you know it won't say something inappropriate?"

It's a fair question, and it deserves a direct answer. Here's exactly what we do.

Layered Content Guidelines

Every story generated by Wisp passes through multiple layers of content filtering before it ever reaches your child.

At the prompt level, we include strict behavioural instructions with every generation request. The AI is explicitly told:

• No violence, even cartoon violence.
• No frightening content, jump scares, or threatening characters.
• No adult themes of any kind.
• No real-world brands, celebrities, or news events.
• Language appropriate for the specified age range at all times.

At the model level, we use AI providers with strong safety fine-tuning. We've tested extensively and chosen models that are reliably conservative with children's content.

At the review level, our team periodically samples generated stories to ensure quality and compliance. When we identify edge cases, we update our guidelines.

Age-Appropriate Personalisation

When you create a child profile, you set an age range. This isn't just metadata — it actively shapes the story.

For a 4–6 year old, stories are short (3–4 scenes), use simple vocabulary, feature familiar settings like forests and farmhouses, and always end with the character safe, happy, and often a little sleepy.

For an 8–10 year old, stories can involve mild mystery, longer narrative arcs, more complex choices, and occasional gentle humour that lands for that age group. Still no peril, still no fear — just more sophistication.

The AI is given the child's age range with every generation, and it adapts accordingly.

No Personal Data in Stories

Here's something worth knowing: we never include your child's last name, location, school, or any specific personal detail in the prompt sent to our AI providers. If you've named a child profile "Lily", that name may appear in a story — but nothing else identifying.

We also don't share child profile data with third parties. The AI providers we use receive only what's needed to generate the story: the age range, story preferences, and a session ID. Nothing more.

What Happens If Something Slips Through?

We'd be lying if we said AI systems are perfect. They're not. That's why we've built a feedback mechanism into every story — you can flag any scene that concerns you, and our team reviews it within 24 hours. We use that feedback to improve our guidelines.

To date, the vast majority of flags we receive are from parents who wanted the story to be even calmer rather than anything inappropriate — which tells us we're calibrated roughly right.

COPPA Compliance

Wisp accounts are created by parents and guardians, not by children. Children do not have independent accounts or the ability to sign up. This keeps us on the right side of COPPA (the Children's Online Privacy Protection Act) and equivalent regulations in the UK and EU.

Child profiles are tied to a parent account, and only the parent can create, edit, or delete them.

You're Always in Control

At any point, you can:

• Review every story your child has had in their Library.
• Delete a story if you want it removed.
• Adjust the child profile's age range or preferences.
• Delete the child profile entirely.
• Close your account and have all associated data permanently removed.

We think trust is built through transparency, not through asking you to just take our word for it. If you ever have a concern about how Wisp handles your family's data, email us at privacy@wispstories.com — we read every message.