Content Moderation Policy

TekkAnimate AI uses a server-side content moderation engine to scan every prompt and uploaded file before any AI generation occurs. This policy describes what content is prohibited, how moderation decisions are made, and how you can appeal a decision.

1. How Moderation Works

When you submit a generation request (animation, video, image, script, scenes, or voice), your prompt is scanned by our moderation engine before any credits are charged or generation begins. The engine checks for prohibited content patterns and public-figure references. If content is blocked, you receive a user-friendly error message and no credits are deducted.

2. Moderation Outcomes

• Allow: The prompt is safe and generation proceeds
• Flag for Review: A public figure is mentioned without clear realistic intent; generation proceeds but the output is logged for review
• Block: The prompt matches a prohibited content rule; generation is refused and no credits are charged

3. Prohibited Content Categories

The following content categories result in an automatic block:

• CSAM / Child Exploitation: Any content sexualizing or exploiting minors
• Non-Consensual Intimate Content: Deepfakes, revenge porn, or "undress" requests
• Human Trafficking: Content promoting the buying, selling, or trafficking of people
• Violent Crime Instructions: Step-by-step guides for murder, torture, bomb-making, or similar
• Terrorism: Terrorist propaganda, recruitment, or attack planning
• Graphic Violence: Gore, beheadings, executions, or snuff content
• Fraud & Identity Theft: Tutorials for stealing identities, credit cards, or bank accounts
• Malware & Cybercrime: Instructions for creating viruses, ransomware, or botnets
• Kidnapping Glorification: Content promoting or teaching kidnapping or hostage-taking
• Public Figure Impersonation: Realistic depictions of identifiable public figures

4. Public Figure Detection

We maintain a curated list of well-known public figures across politics, entertainment, sports, and social media. When a prompt references one of these figures, the engine checks for realistic-depiction keywords (e.g., "photorealistic," "real," "lookalike," "impersonate," "deepfake"). If realistic intent is detected, the prompt is blocked. If a public figure is mentioned without realistic intent, the prompt is flagged for review but generation proceeds.

5. File Upload Validation

Uploaded reference images are validated for file type (JPG, PNG, WebP, GIF) and file size (maximum 10MB) before processing. Files that fail validation are rejected with a user-friendly error message.

6. Moderation Logging

Every moderation decision is logged to a secure ModerationLog entity, including:

• User ID and email
• Project ID and Job ID
• The full prompt that was scanned
• The generation type (animation, video, image, script, scenes, voice)
• Risk category (safe, review, block)
• Decision (allow, flag, block)
• Matched rules and blocked reason
• Public figure detection result
• Timestamp

Moderation logs are accessible only to platform administrators and are used to improve our safety systems and investigate violations.

7. Appeals

If you believe your content was incorrectly blocked, you can appeal through our Contact page. Provide your project details and the prompt that was blocked. Our team will review the decision and respond within 48 hours. Approved appeals will allow you to regenerate the content without additional credit charges.

8. Consequences of Violations

Repeated or severe violations may result in:

• Content blocking and credit forfeiture for the attempted generation
• Account warnings and temporary restrictions
• Permanent account suspension
• Referral to law enforcement for illegal content