How AI is used
Text, such as outlines, chapters and comic scripts, is generated by Anthropic Claude. Images, such as character portraits, covers and comic panels, are generated by Cloudflare Workers AI using FLUX models. The service builds each prompt from your project and your Story Bible, runs the generation, and saves the result as a version you can keep, replace or delete.
Labelling and provenance
Every image the service generates is clearly marked as AI-generated:
- The file carries embedded provenance metadata. We set the IPTC digital source type to
trainedAlgorithmicMedia, the standard value for media made by a generative model. - The image is shown with a visible AI label in the reader.
You must not remove or hide these labels or the embedded metadata, and you must not present AI-generated media as if it were made without AI.
What you must not create
- No realistic depictions of real, identifiable people, and no deepfakes.
- No forged documents, certificates or official records meant to deceive.
- Fiction only. Characters must be invented, not real people.
The full list of content rules is in the Acceptable Use Policy, and the service applies the same rules when it generates or checks content.
Character consistency is best effort
Image models do not remember the face of a character between generations. Attaching a reference portrait helps a character stay recognisable from one image to the next, but we do not and cannot guarantee a perfect match. The service tells you which model it used and keeps every version so you can choose the best one.
Your responsibility for published works
You decide what to publish. You are responsible for making sure your published work follows this policy, the Acceptable Use Policy and the law, and that it does not infringe the rights of anyone else. Keep the AI labelling intact when you share or export your work.
How we respond to unlawful synthetic content
We act on unlawful synthetic content within these timelines:
- Lawful government or court orders: within 3 hours.
- Specified complaints about intimate imagery or impersonation: within 2 hours of a valid complaint.
- Other grievances: acknowledged within 24 hours and disposed within 7 days.
Report content from the Grievance Officer page.
Periodic reminders
We remind users of these rules at least once every three months, in line with the information technology rules as amended in February 2026.
The rules we follow
This policy follows the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, as amended by the 2026 amendment that took effect on 20 February 2026. In plain terms, the amended rules require us to:
- Label synthetically generated information and keep provenance metadata with it.
- Remove content within 3 hours of a lawful government or court order.
- Remove specified intimate-imagery or impersonation content within 2 hours of a valid complaint.
- Dispose of grievances within 7 days.
- Remind users of the applicable rules periodically.
This is a plain-language summary, not legal advice. The rules themselves govern.