AUP

Pawmory (“Platform”, “we”) lets you turn a pet photo into a short video. This Acceptable Use Policy (“Policy”) applies to all content uploaded, submitted, generated, stored, or distributed through the Service. It sets our content standards, moderation, reporting, and enforcement. Using the Service means you agree to this Policy, the Terms of Service, and the Privacy Policy.

1. Scope

This Policy covers:

• pet photos and other images you upload;
• scene notes, captions, and other prompts you submit;
• short videos we generate, including picture, motion, and sound;
• account information and messages you send to support.

Video generation uses MiniMax (via MuAPI). You must also follow MuAPI and MiniMax acceptable-use rules. We do not offer uncensored, “spicy”, or open-weights endpoints.

2. Content standards and prohibited categories

The Platform prohibits generating the following six categories of content:

• Pornography / NSFW content.
• Violence / gore.
• Hate speech.
• Child-unsafe content (CSAM).
• Deepfakes / impersonation of real people.
• Content infringing copyright or trademarks.

The Service is only for memory videos from pet photos you have the right to use. None of the six categories is allowed. We do not offer an age-gated adult mode.

2.1 Absolutely prohibited — zero tolerance

The following will be actioned immediately and may be reported to authorities:

• Child sexual abuse material (CSAM) or any sexual content involving minors;
• content that promotes or facilitates terrorism, extremism, or mass violence;
• technical instructions for weapons of mass destruction;
• content inciting genocide or ethnic hatred;
• non-consensual deepfake sexual content targeting real individuals;
• any other content prohibited by applicable law, including the laws of the People’s Republic of China.

2.2 Extra limits for this product

Pawmory is a pet-memory product. We do not offer adult, political, or professional-advice use cases. In addition to the six categories above:

• nudity, pornography, or any NSFW content, including claims of artistic or educational purpose;
• graphic gore, torture, or extreme violence;
• political or current-affairs deepfakes that could mislead the public;
• presenting Output as medical, legal, or financial advice.

2.3 Generally prohibited conduct

• generating or spreading misinformation;
• impersonating real people, brands, or organizations;
• infringing intellectual property, publicity, or privacy rights, including using someone else’s photos or protected characters;
• passing AI-generated content off as a real recording or human-made work;
• bulk-generating spam or deceptive commercial content;
• jailbreak prompts designed to bypass safety filters;
• violating MuAPI’s acceptable use rules, including pornography, violence, hate, deepfakes, and harmful synthetic media.

3. Platform content moderation（平台内容审核机制）

We operate a multi-layer system of automated review, human review, and periodic audits. Generated videos are private to the account holder by default and are not posted to a public gallery. After a user report we follow the review flow, review standards, response timelines, and result-notification rules below.

3.1 Automated review (real-time)

All generation requests and outputs are scanned by safety filters before delivery.

• Input filtering: before generation we scan the user’s scene note with our keyword/phrase filter, then call Waffo Prompt Sift (scan-prompt, semantic mode enforce). We continue only when the verdict is allow. A block is refused immediately with no credits held. review or a degraded safety service does not generate. Prompt Sift does not retain prompt text; we store the verdict and a hash of the prompt, not the raw text, in our scan log.
• Output filtering: requests go to MiniMax via MuAPI. MiniMax / MuAPI apply their built-in safety filters to inputs and outputs. Requests that trip those filters are refused and no video is delivered.
• Usage pattern analysis: we watch for unusually high failure rates, repeated prohibited prompts, and abusive sign-ups.
• Safety model used: Waffo Prompt Sift, Pawmory’s keyword / phrase filter, plus MiniMax / MuAPI built-in safety filters.

3.2 Human review

• Auto-blocked or flagged content enters a human review queue.
• After a user report, attended support reviews the photo, prompt, and output against Section 2 and the L1–L4 standards in 3.4.
• Review is handled through the attended inbox support@pawmory.org. We do not offer a 24/7 hotline.

3.3 Periodic audits

At least weekly we randomly sample recent generations and refusals to improve the blocklist and scene templates. We do not currently use an independent third-party auditor.

3.4 Content severity levels (review standards)

3.5 After a report: review, timelines, and notification

Processing workflow: report received (email acknowledgement) → triage (classify L1–L4) → review (automated records + human) → decision → enforcement → notification of the result.

• How we review: we check the reported photo, prompt, video, and account record against the six prohibited categories in Section 2 and the L1–L4 standards in Section 3.4.
• Review standards: Section 2 decides whether content violates this Policy; Section 3.4 decides priority and the enforcement level.
• Timelines: L1 first response within 24 hours, resolution within 48 hours; L2 within 24 hours / 3 business days; L3 within 3 / 7 business days; L4 within 5 / 15 business days. See Section 6.
• Result notification: we email the address that filed the report, stating whether the report is upheld and what we did (warning, removal, feature limit, suspension, or account closure). If we cannot reach the reporter we still enforce and keep a record.

4. Reporting

4.1 When to report

• generated or uploaded content that violates this Policy, especially the six prohibited categories in Section 2;
• critical issues involving child safety or terrorism;
• AI impersonation, use of your likeness without consent, or infringement of your copyright or trademark;
• attempts to bypass safety filters.

4.2 Principles

• Confidentiality: we do not disclose a reporter’s identity to the reported party except where law requires;
• Impartiality: reports are reviewed on the materials;
• Anti-abuse: malicious or repeat false reports are logged and may restrict the account.

5. Reporting channel

Email support at support@pawmory.org. Use the subject “Content report” and include where you can:

• your account email;
• the generation or work link / ID, if you have one;
• a short note on which prohibited category applies.

This is the only attended channel. Do not use unpublished addresses.

6. Response timelines

If we need more time, we will notify the reporter and say why. Processing workflow: report received (email acknowledgement) → triage (classify L1–L4) → review (automated + human) → decision → enforcement → notification of the result.

7. Enforcement

8. Appeals

If you disagree with an enforcement decision, email support@pawmory.org within 10 business days of notice. Use the subject “Content appeal”, explain why, and include the original task information. We will re-examine the original materials and reply within 10 business days.

9. Policy updates

Material changes will be emailed at least 15 days in advance and the date on this page will be updated. Continued use is acceptance. This Policy is not legal advice.

10. Contact

Safety, reports, and appeals all go to support at support@pawmory.org. Pawmory · https://www.pawmory.org

Last updated: 16 September 2026 · Version 1.3