1. Content Standards & Prohibited Categories
Malfoid is an anime and fandom creative studio designed for general audiences. You may not use Malfoid to generate or attempt to generate any content falling into the following six prohibited categories:
- Pornographic / NSFW content: Sexually explicit imagery, nudity, sexualized depictions, fetish content, or any content whose primary purpose is sexual gratification. Malfoid produces stylized anime character art for general audiences — adult, erotic, or NSFW content of any kind is strictly forbidden.
- Violence / gore: Graphic violence, blood, gore, mutilation, decapitation, torture, self-harm, or content that glorifies or incites physical injury or cruelty against people or animals.
- Hate speech & harassment: Content that attacks, demeans, defames, or dehumanizes individuals or groups based on race, ethnicity, national origin, religion, gender, gender identity, sexual orientation, disability, or immigration status, including hate symbols and slurs.
- Child-unsafe content: Any sexualization of minors, child sexual abuse material (CSAM), or content depicting minors in a suggestive or sexually exploitative context. This category is subject to zero tolerance (see below).
- Deepfakes / impersonation: Photorealistic or identifiable depictions of real, living people (including public officials, politicians, and celebrities) without explicit consent; impersonation of real individuals, brands, or organizations; or content presented as authentic evidence of real-world events.
- Copyright / trademark infringement: Prompts intended to directly reproduce third-party copyrighted characters, commercial logos, registered trademarks, or proprietary franchise assets. Names of commercial franchise characters and rights holders are blocked by our automated input filter. Malfoid is an independent fan-made concept and parody meme not affiliated with or endorsed by any rights holder (see our Disclaimer page).
Zero-Tolerance Content
The following violations result in immediate permanent bans, instant removal of all generated content, and where required by law, formal reports to relevant law enforcement and child-protection organizations (such as the National Center for Missing & Exploited Children):
- Any sexual content involving minors, including CSAM or any attempt to produce it (all attempts are logged and reported).
- Terrorist, violent extremist, or mass-casualty content.
- Non-consensual sexualized or intimate depictions of real people.
In addition to the categories above, you may not: attempt to bypass, jailbreak, or circumvent our prompt filters or the upstream model's safety systems; generate content that violates applicable laws; use the service for automated scraping, bulk abuse, or harassment; or knowingly submit fraudulent abuse reports.
2. Content Moderation Architecture
We maintain community-friendly generations through a layered defense architecture:
- Automated real-time input filtering: Every user prompt and trait parameter is evaluated in real time against a multi-tier keyword blacklist before any GPU compute is dispatched. The filter detects obfuscation, leetspeak substitutions, and spacing tricks across NSFW, violence, hate speech, child safety, real public figures, and third-party trademarks. User prompts are also strictly limited to a maximum of 500 characters.
- Safety-reviewed prompt templating: Generations are synthesized from structured, curated prompt templates and predefined trait presets. User input functions strictly as an aesthetic supplement rather than unbounded control over raw model weights.
- Provider-side model safety & recognized AI architectures: Generations run through enterprise cloud inference clusters via Cloudflare AI Gateway hosting the recognized foundation model: Google Nano Banana (
google/nano-banana-2-lite) developed by Google. This foundation model operates with safety-evaluated neural weights and built-in inference classifiers that detect and suppress policy-violating imagery at the architecture level. - Mode restrictions (Text-to-Image only): To safeguard personal privacy and prevent non-consensual imagery, Malfoid operates strictly in text-to-image mode. Photo uploads and image-to-image transformations are disabled, preventing any ingestion or storage of personal portrait photos.
- Human review: Prompts flagged by our system or reported by users are reviewed by trained administrators before permanent account actions are finalized.
- Periodic safety audits: We continuously review blocked prompts, monitor evasion techniques, and expand our keyword blacklists to address emerging risks.
Severity Levels & Response Targets
| Level | Category Examples | Automated & Human Action | Target Window |
|---|---|---|---|
| L1 — Critical | CSAM, terrorism, non-consensual intimate imagery | Instant generation block; permanent account termination; external law enforcement referral | Within 24 hours |
| L2 — Severe | Graphic gore, hate speech, deepfakes / impersonation | Content deletion; temporary account suspension pending review | Within 3 business days |
| L3 — General | NSFW attempts, trademark infringement, filter evasion | Prompt blocked; content removal; formal written warning | Within 7 business days |
| L4 — Minor | Borderline prompt phrasing, excessive prompt length | Input validation error; prompt adjustment request | Within 15 business days |
3. Reporting Channels
If you encounter generated content on Malfoid that violates this policy, please notify us immediately through our designated reporting channel:
Email: [email protected]
Subject Line: Content Report — AIGC Violation
What to Include:
- The image file, image URL, or screenshot of the generation
- Generation timestamp or task ID if displayed in your studio session
- A concise explanation of the alleged violation category
We handle all reports confidentially and do not disclose the identity of reporters to reported users. We maintain a strict non-retaliation policy for good-faith reports. Knowingly fraudulent or abusive reports may result in enforcement action against the submitter.
4. Enforcement Actions
Confirmed violations trigger a graduated enforcement response based on severity and prior history:
- Warning: Formal notice issued for first-time minor or borderline infractions.
- Content removal: Confirmed-violating images are purged from our servers and generation history.
- Feature restrictions: Rate limiting, reduced generation quotas, or temporary revocation of style modes for repeat infractions.
- Credit forfeiture: No refunds or credit compensation are provided for credits expended on generations removed or blocked due to confirmed policy violations.
- Account suspension: Temporary suspension of account access pending administrative review for severe violations.
- Permanent ban: Immediate termination of user account and permanent revocation of platform access for zero-tolerance content or repeated severe violations; all remaining credits are forfeited.
- Legal referral: Where content involves unlawful conduct (such as child sexual abuse material or actionable threats), we refer the matter and relevant session logs to law enforcement agencies as required by statute.
Violations accumulate across accounts: repeated minor infractions escalate to higher penalty tiers regardless of individual severity.
5. Appeals Process
If you believe an enforcement action or filter block was executed in error, you may submit a formal appeal:
Email [email protected] within 30 calendar days of the enforcement action with the subject line AUP Appeal — [Your Email / Account]. Appeals are reviewed by team members who were not involved in the initial enforcement decision.
6. Updates to This Policy
We may periodically revise this Acceptable Use Policy as our generator capabilities, community standards, and upstream AI models evolve. Any changes take effect immediately upon posting to this page.
7. Contact Information
For questions or inquiries regarding our AIGC policies, contact:
Email: [email protected]
See also our AI Art Studio, About Malfoid, and Disclaimer & DMCA Policy.