Effective date: 2 July 2026

Content Moderation Policy

SafeVoice handles sensitive survivor-generated content. This policy explains how we review and moderate content on our platform.

1. Human Review

All submitted reports are reviewed by trained WGHEN caseworkers before routing to NGOs, legal aid, or authorities. AI assists with triage but does not make final decisions.

2. AI Output Moderation

AI chat responses are filtered for harmful content, crisis escalation, and prompt-injection attempts. Users can flag problematic AI output via Report Abuse.

3. Evidence Review

Uploaded images and audio are accessed only by authorized caseworkers. EXIF metadata is stripped where possible. Evidence is stored in encrypted private storage.

4. Removal Actions

We may remove or quarantine content that is spam, malicious, unrelated to GBV support, or violates our Acceptable Use Policy.

5. Appeals

If you believe your content was wrongly removed, contact safety@safevoice.ng within 14 days.