BiasHigh
Bias in AI Deepfake Detection Undermines Election Security in Global South
- Occurred
- Sep 2, 2024
- Published
- Sep 2, 2024
- Source
- AI Incident Database
Summary
AI deepfake detection tools are reportedly failing voters in the Global South due to biases in their training data. These tools, which prioritize English language and Western faces, show reduced accuracy when detecting manipulated content from non-Western regions. As a result of this detection gap, election integrity faces threats from and the amplification of misinformation, which leaves journalists and researchers with inadequate resources to combat the issue.
Developers
- Unknown deepfake detection technology developers
- True Media
- Reality Defender
Deployers
- Unknown deepfake detection technology developers
- True Media
- Reality Defender
Harmed parties
- Global South Citizens
- Political researchers
- Global South local fact-checkers
- Non-native English speakers
- Global South journalists
- Civil society organizations in developing countries
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: AI detection tools exhibit demographic bias affecting election security in Global South.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8