AI Content Policy Library

Guidelines for creating prompts and content that comply with AI platform policies

Version: 1.0 | Updated: Feb 2026 Author: Nguyen Ngoc Tuan | fb/khongphaituan

---

Table of Contents

1. Introduction
2. Document Structure
3. Supported Platforms
4. Usage Workflow
5. Strictness Levels
6. Key Principles
7. Changelog

---

Introduction

When developing prompts, AI workflows, or creating AI-generated content, understanding each platform's Community Guidelines and Usage Policies is crucial.

This library compiles and analyzes content policies from popular AI platforms to help you:

• Avoid policy violations and blocks
• Create safe, compliant content
• Understand each platform's limitations
• Optimize prompts for specific use cases

---

Document Structure

File	Description	When to Use
01-PROHIBITED-CONTENT-UNIVERSAL.md	Content prohibited on ALL platforms	Before writing any prompt
02-SENSITIVE-TOPICS-GUIDE.md	Sensitive topics and safe handling	Dealing with sensitive content
03-PLATFORM-SPECIFIC-POLICIES.md	Detailed policy comparison by platform	Multi-platform deployment
04-COPYRIGHT-IP-GUIDELINES.md	Copyright and IP regulations	IP-related content
05-HISTORICAL-POLITICAL-CONTENT.md	Historical, political, public figures	Historical/political content
06-PROMPT-COMPLIANCE-CHECKLIST.md	Pre-deployment prompt checklist	Final compliance check
07-COMMON-TRIGGERS-KEYWORDS.md	Keywords and patterns often blocked	Debugging rejected prompts
08-WORKAROUNDS-BEST-PRACTICES.md	Safe content rewriting techniques	Content rewriting

---

Supported Platforms

AI Generation Platforms

• Google Gemini / Vertex AI - AI Studio & Enterprise API
• OpenAI ChatGPT / GPT API - Consumer & Developer
• Anthropic Claude - Consumer & API

Social Media (AI-generated content)

• YouTube - AI-generated content policies
• TikTok - AI content guidelines

---

Usage Workflow

┌─────────────────────────────────────────────────────────────────┐
│  LIBRARY USAGE WORKFLOW                                         │
└─────────────────────────────────────────────────────────────────┘

    ┌──────────────┐
    │    Start     │
    │   writing    │
    │   prompt     │
    └──────┬───────┘
           │
           ▼
    ┌──────────────────────────────────────┐
    │ 1. Check 01-PROHIBITED-CONTENT       │
    │    → Is content prohibited?          │
    └──────────────┬───────────────────────┘
                   │
           ┌───────┴───────┐
           │               │
       Prohibited        OK ✓
           ✗               │
           │               │
     ┌─────▼─────┐   ┌─────▼──────────────────┐
     │   STOP    │   │ 2. Check 02-SENSITIVE  │
     │   Do not  │   │    → Sensitive topic?  │
     │  proceed  │   └─────┬──────────────────┘
     └───────────┘         │
                   ┌───────┴───────┐
                   │               │
                  Yes             No
                   │               │
     ┌─────────────▼───────┐       │
     │ 3. Apply guidelines │       │
     │    from guide       │       │
     └─────────────┬───────┘       │
                   │               │
                   └───────┬───────┘
                           │
                           ▼
             ┌─────────────────────────────┐
             │ 4. Check 03-PLATFORM        │
             │    → Which platform?        │
             └─────────────┬───────────────┘
                           │
                           ▼
             ┌─────────────────────────────┐
             │ 5. Final: 06-CHECKLIST      │
             │    → Compliance check       │
             └─────────────┬───────────────┘
                           │
                           ▼
                    ┌──────────────┐
                    │   Complete   │
                    │  ✓ Safe      │
                    │    Prompt    │
                    └──────────────┘

Quick Steps

1. Before writing: Check 01-PROHIBITED-CONTENT-UNIVERSAL.md
2. Sensitive topics: See 02-SENSITIVE-TOPICS-GUIDE.md
3. Multi-platform: Compare in 03-PLATFORM-SPECIFIC-POLICIES.md
4. Final check: Use 06-PROMPT-COMPLIANCE-CHECKLIST.md
5. If blocked: Debug with 07-COMMON-TRIGGERS-KEYWORDS.md

---

Strictness Levels

Most Strict ◄─────────────────────────────────────────► Most Flexible

┌──────────────┬──────────────┬──────────────┬──────────────┬──────────────┐
│  Vertex AI   │  OpenAI API  │  Claude API  │  AI Studio   │   ChatGPT    │
│ (Enterprise) │   (Paid)     │    (API)     │   (Free)     │   Consumer   │
└──────────────┴──────────────┴──────────────┴──────────────┴──────────────┘
      ★★★★★          ★★★★           ★★★            ★★              ★

Note: Vertex AI (Google Cloud) has enterprise-level content filters significantly stricter than AI Studio (consumer).

---

Key Principles

Hardcoded Restrictions (Cannot bypass)

Some content is absolutely prohibited at the model level with no possible workaround:

Type	Description
CSAM	Child Sexual Abuse Material
CBRN	Weapons of mass destruction instructions
Malware	Malware and cyberattack tools
Deepfake	Deepfakes of real people for fraud

Context-Dependent

Some content may be acceptable with appropriate context:

Context	Example
Educational	Historical or scientific explanations
Historical	Historical documentation, archives
Creative Fiction	Novels, films (with disclaimer)
Professional Research	Academic research

---

Changelog

Date	Change
Feb 2026	Initial version
Feb 2026	Restructured, added author info, English translation

---

Contributing

For suggestions or contributions, please contact:

• Facebook: fb/khongphaituan
• Author: Nguyen Ngoc Tuan

---

This documentation is created for educational purposes and responsible AI usage guidance.