Our Commitment to Safety
At AiEnglish, we're committed to a safe, educational environment. Our upgraded content safety uses OpenAI's Moderation API to provide accurate AI-powered analysis while maintaining an educational focus.
AI-Powered Moderation
All conversations are analyzed in real-time by the OpenAI Moderation API. The AI provides confidence scores, enabling transparent safety decisions with minimal blocking.
How It Works
- Real-time AI Analysis: Conversations are evaluated as they happen
- Confidence Scoring: Content is blocked only when AI confidence is high (typically > 70%)
- Advanced Categorization: AI identifies content categories with high accuracy
- No Keyword Lists: Uses advanced AI analysis instead of simple pattern matching
Content Categories We Block
🚫 Violence
Physical harm, weapons, threats, war-related content
🔞 Adult Content
Sexual material, explicit content, inappropriate intimacy
Safety Levels & Responses
Our AI classifies content and responds appropriately, prioritizing education with minimal blocking:
✅ Safe
Content is appropriate for language learning. No action needed.
⚠️ Warning
Borderline content may receive a reminder while remaining visible for educational context.
🚫 Blocked
Only violence and adult content are blocked when AI confidence is high. Users are redirected to appropriate topics.
🔴 Critical
Content is harmful or dangerous. Immediate action taken, and user may be temporarily restricted.
User Controls & Customization
You have control over AI safety settings and can customize your experience:
Safety Settings
- Safety Toggle: Enable or disable AI content safety
- Blocked Categories: Violence and adult content blocked by default
- Educational Focus: Educational content is always allowed
Data Privacy & Safety
Your privacy is protected even while maintaining safety:
Real-Time Analysis Only
All content safety analysis happens in real-time using OpenAI's Moderation API:
- Real-Time Processing: Conversations are analyzed as they happen
- Immediate Discard: Content is analyzed and immediately discarded—we don't store conversation content
- No Content Storage: Your conversation text is never stored on our servers
- No Moderation Records: We don't keep records of moderation actions or decisions
What We Don't Store
We don't store your conversation content on our servers. Content is analyzed in real-time and immediately discarded. For full details, please see our Privacy Policy.
Analytics & Privacy
- Aggregated Analytics Only: We send only aggregated analytics events (without any conversation content) to Firebase for service improvement
- No Content in Analytics: Analytics events contain no conversation text or user content
- Automatic Filtering: All content filtering is automatic—there is no user reporting system
Getting Help
If you have questions about content safety:
Our Safety Promise
We promise to:
- Maintain a safe, educational environment for all users
- Protect your privacy while ensuring safety
- Respond quickly to safety concerns
- Continuously improve our safety systems
- Be transparent about our safety practices
- Respect your right to learn in a safe environment
Safety is Our Priority
Every feature in AiEnglish is designed with safety in mind. From our AI conversations to our content filtering, we work tirelessly to ensure that your language learning journey is both effective and safe.