1. Scope
This Acceptable Use Policy ("Policy") governs all content generated, uploaded, or processed through PDFChat's AI-powered PDF services. This includes AI-generated answers, summaries, extracted text, and any other output produced by our Service. "Content" includes user-submitted prompts, uploaded PDF documents, and AI-generated responses.
2. Content Standards and Prohibited Categories
PDFChat explicitly prohibits using this Service to generate, upload, or distribute the following six categories of content:
- Adult / NSFW Content — Sexually explicit or pornographic material of any kind
- Violence / Gore — Content depicting or promoting physical harm, injury, or violent acts
- Hate Speech — Content that promotes discrimination, hatred, or violence against individuals or groups based on protected characteristics
- Child Safety (CSAM) — Any content sexualizing minors or depicting child exploitation in any form
- Deepfakes / Impersonation — Synthetic media impersonating real individuals or used to deceive viewers about source or authenticity
- Copyright / Trademark Infringement — Content reproducing copyrighted works or trademarks without authorization
2.1 Absolute Prohibitions — Zero Tolerance
The following violations result in immediate account termination and mandatory reporting to authorities:
- Child sexual abuse material (CSAM) or any sexualization of minors
- Terrorism-promoting content or instructions for creating weapons of mass destruction
- Content advocating genocide or racial hatred
- Non-consensual deepfake sexual content
- Content that facilitates serious crimes against persons
- Adult / sexually explicit content of any kind
- Violence or graphic gore content
- Hate speech or content promoting discrimination
2.2 General Prohibited Conduct
- Generating or spreading misinformation, rumors, or false content
- Impersonating real persons, brands, or institutions
- Infringing intellectual property, portrait rights, or privacy rights
- Presenting AI-generated content as human-created content
- Bulk-generating spam or fraudulent content
- Using jailbreak prompts or attempts to bypass our content safety filters
- Using AI-generated content to impersonate professionals (doctors, lawyers, financial advisors) in contexts where it could cause harm
- Using Service inputs, outputs, or model responses to train, fine-tune, or develop any competing AI model
- Systematically scraping or extracting model outputs beyond personal authorized use
3. Content Moderation Mechanism
PDFChat employs a multi-layer moderation system: automated real-time scanning + human review + periodic audits.
3.1 Automated Moderation (Real-Time)
All prompts and AI-generated responses are scanned in real-time by our safety filters:
- Input Filtering: Keyword, semantic, and intent analysis of user-submitted prompts
- Output Filtering: Real-time safety assessment of AI-generated responses before they are presented to the user
- Usage Pattern Analysis: Monitoring for abnormal API call behavior indicative of abuse
- Moderation Tools: Every user prompt is screened through the Waffo Pancake content-safety scanning API (Prompt Sift) before it reaches the generation model, and responses pass through the built-in safety systems of Alibaba Cloud Qwen. Input and output are both checked before responses are returned.
3.2 Human Review
- Automatically flagged content enters a human review queue
- User-reported content is investigated by our Trust & Safety team
- Reviewers complete content safety training and sign confidentiality agreements
3.3 Periodic Audits
- Random sampling review is conducted monthly
- Automated moderation system accuracy is evaluated and optimized regularly
- Audit results are reviewed by the Trust & Safety team and inform system improvements
3.4 Content Severity Classification
| Level | Description | Review | Response |
|---|
| L1 — Critical | CSAM, terrorism, genocide advocacy | Auto-block + human confirmation | Immediate account ban + report authorities |
| L2 — High Risk | NSFW, deepfakes, hate speech | Auto-flag + priority review | Content removal + account suspension |
| L3 — Medium Risk | Copyright infringement, misleading AI content | Human review | Content removal + warning |
| L4 — Low Risk | Minor violations, first-time offenses | Report-triggered | Warning + content modification request |
4. Reporting Mechanism
4.1 When to Report
- Content that violates this Policy's content standards
- Content involving child safety, terrorism, or other serious violations
- Discovery of AI impersonation or infringement of your own rights
4.2 Reporting Principles
- Confidentiality: Reporter identity is strictly protected
- Neutrality: All reports undergo independent, objective review
- Anti-Abuse: Malicious or false reports are documented and may result in action against the reporter
5. Reporting Channels
For urgent matters involving imminent harm, contact us immediately via email. All reports are treated confidentially.
6. Response and Resolution Times
| Violation Level | Initial Response | Resolution |
|---|
| L1 — Critical | Within 2 hours | Within 24 hours |
| L2 — High Risk | Within 24 hours | Within 3 business days |
| L3 — Medium Risk | Within 3 business days | Within 7 business days |
| L4 — Low Risk | Within 5 business days | Within 15 business days |
Processing flow: Report received (auto-confirmation) → Initial severity classification (L1–L4) → Content review (automated + human) → Enforcement decision → Enforcement execution → Result notification. If extended processing time is needed, we will proactively notify the reporter with an explanation.
7. Enforcement Actions
| Action | Description | Applied When |
|---|
| 1. Content Warning | Written notice requesting removal or modification | First minor violation |
| 2. Content Removal | Violating content is taken down | Confirmed L3–L4 |
| 3. Feature Limitation | API access or generation features restricted | Multiple minor violations or single L3 |
| 4. Account Suspension | Temporary suspension pending review | L2 or accumulated violations |
| 5. Permanent Ban | Permanently terminated access | L1 or multiple L2 violations |
| 6. Legal Action | Report to law enforcement and cooperate with investigations | Illegal content suspected |
8. Appeal Process
If you believe your content was incorrectly flagged or your account was wrongly suspended, you may file an appeal within 30 days of receiving the notification:
- Appeal Email: appeal@pdfchat.tech
- Review Timeline: 10 business days for independent re-evaluation by a reviewer not involved in the original decision
9. Transparency and Policy Updates
- We publish a content safety transparency report periodically
- Material changes to this Policy will be communicated at least 15 days in advance via email or website notice
- Continued use of the Platform after changes constitutes acceptance of the updated Policy