Sign In
Terms & Policies

Content Moderation Policy

Last Modified: April 20, 2026

This Content Moderation Policy is published by SugarGlitch Ltd, a Cyprus private limited company ("SugarGlitch," "we," "our," or "us"), and is incorporated into our Terms of Service.

This Policy describes our approach to moderating Content on the Services. It works alongside our Prohibited Content Policy (which describes what is not allowed), our Content Removal Policy (which describes when and how Content is removed), our Pre-Screening and Post-Screening Policy (which describes our review processes in more detail), and our Complaint Policy (which describes how to raise concerns).

1. Our Approach

SugarGlitch is an AI generation platform. We support a wide range of creative expression, including adult content for users 18 and older, but we do not allow the Services to be used for harm or unlawful purposes. Our moderation approach is guided by a few principles:

  • Safety over scale. We treat the prohibited categories in our Prohibited Content Policy — particularly child safety and real-person sexual exploitation — as non-negotiable, even where enforcement creates friction or false positives.
  • Proportionality. For lower-stakes violations, our response is proportionate to the severity, frequency, and intent of the violation. We do not treat a single edge-case prompt the same as a pattern of deliberate abuse.
  • Honesty about limitations. No moderation system catches everything. We are open about what we can and cannot reliably detect.
  • Respect for legitimate creative use. We do not moderate adult content as a category. Our enforcement targets specific harms — not the genre.

2. What We Moderate

This Policy applies to all Content on the Services, including:

  • AI-generated outputs (text, images, audio, video, and other Generations)
  • Prompts and inputs submitted by users
  • Characters and their metadata (names, descriptions, traits, voices, instructions)
  • Usernames, display names, profile fields, and avatars
  • User communications visible on the Services, including comments, public Character pages, and shared content
  • Behavior and patterns of activity on the Services

3. Moderation Methods

We use a combination of automated systems, user reports, human review, and proactive audits. These are described in more detail in our Pre-Screening and Post-Screening Policy.

Automated systems. We use classifiers, content filters, and rule-based systems to detect and prevent violations of our Prohibited Content Policy. This includes:

  • Prompt and input filtering, including pattern detection for known abuse vectors and jailbreak techniques
  • Output classification on Generations, including for SFW/NSFW categorization and for prohibited categories
  • Detection of attempts to depict minors in sexual contexts, with strict thresholds and a low tolerance for false negatives in this category
  • Account-level signals such as repeat-violation patterns and coordinated activity

Human review. We use human review for content and accounts that are flagged by our automated systems, reported by users, escalated for legal or safety reasons, or otherwise identified as warranting individual judgment. Human review is also applied to consequential enforcement decisions (such as account terminations) where practicable.

User reports. Reports from users are an important moderation input. We aim to acknowledge reports promptly and to respond within a reasonable time — typically within 7 business days, sooner for urgent or high-severity reports.

Proactive audits. We may proactively review trending Characters, popular prompts, frequently-flagged accounts, and other patterns of platform activity to identify systemic issues and adjust our systems accordingly.

4. Privacy and Moderation

Some moderation occurs on Content that is otherwise private to you, including prompts you submit and chats with Characters. This is necessary to enforce our Prohibited Content Policy — particularly with respect to child safety and other zero-tolerance categories. We take care to limit such review to what is needed for moderation purposes, and we handle any personal information involved in accordance with our Privacy Policy.

We do not review private content for purposes unrelated to safety, security, fraud prevention, legal compliance, or improving the Services as described in our Privacy Policy.

5. Limitations of Moderation

Automated systems and human review are both imperfect:

  • Automated systems may flag Content that does not actually violate our policies (false positives) and may miss Content that does (false negatives). This is particularly true for context-sensitive judgments such as parody, fictional roleplay, or transformative use.
  • Human review is constrained by volume, language coverage, and the inherently subjective nature of some moderation judgments.
  • AI-generated outputs are inherently variable — the same prompt can produce different outputs, and Characters may produce material their creator did not anticipate.

We continuously work to improve detection accuracy, but we ask users to understand that moderation is not a perfect filter. When a system makes a wrong call, the appeals process described in our Content Removal Policy is the remedy.

6. Reporting Content

If you encounter Content on the Services that you believe violates our Prohibited Content Policy or other rules, please report it to [email protected] with the subject line "Content Report." Include a description of the Content, where it appears, and why you believe it violates our policies. See our Complaint Policy for more on the reporting process.

For suspected child sexual abuse material, you may also report directly to:

  • The U.S. National Center for Missing & Exploited Children's CyberTipline at https://report.cybertip.org
  • The relevant authority in your jurisdiction (in the EU, INHOPE-affiliated hotlines)

7. Consequences for Violations

Depending on the severity, frequency, and circumstances of a violation, our response may include:

  • Removal or restriction of the offending Content
  • Editing or deletion of Character metadata, usernames, or other profile content
  • Restriction of specific features (for example, NSFW access or generation features)
  • Warning, temporary suspension, or permanent termination of the account
  • Forfeiture of unused Credits or subscription benefits
  • Reporting to law enforcement where required or appropriate

We take immediate action without prior notice for the highest-severity categories, including child sexual abuse material, real-person sexual content, credible threats, and trafficking content. Procedures and appeal rights are described in the Content Removal Policy.

8. User Responsibility

You are responsible for your prompts, Characters, Generations, and other Content. Attempting to evade our moderation systems — for example, through prompt manipulation, jailbreak techniques, the use of euphemisms or substituted characters to disguise prohibited Content, or the re-creation of removed Content — is itself a violation of our Terms of Service and may result in termination.

You are also responsible for evaluating any Generation before sharing or relying on it. Outputs are not pre-approved by SugarGlitch, and the fact that a Generation was produced by our systems does not mean it is appropriate for any particular use.

9. Continuous Improvement

We update our moderation systems, filters, and procedures over time as we identify new patterns of abuse, new model behaviors, and new legal or industry developments. We may also adjust this Policy from time to time to reflect those changes.

10. Contact Us

For questions about this Policy or our moderation practices:

SugarGlitch Ltd
[REGISTERED ADDRESS — TBD]
Cyprus
Email: [email protected]