Deepfake Detection & Media Authentication

Source: https://github.com/dirnbauer/webconsulting-skills

Comprehensive framework for detecting synthetic media, analyzing manipulation artifacts, and establishing media provenance in the post-empirical era.

Key Insight: Traditional detection methods (PRNU, IGH, DQ) are like fingerprints—helpful, but disputable. Cryptographic provenance (C2PA) is like a DNA match—mathematically certain (collision probability 2⁻²⁵⁶).

When to Use

• Verifying authenticity of images or videos before publication
• Detecting AI-generated or manipulated media (deepfakes, face swaps, synthetic voices)
• Forensic analysis of suspicious media for legal or journalistic purposes
• Implementing automated media authentication pipelines
• Establishing content provenance and chain of custody
• Countering disinformation campaigns and Advanced Persistent Manipulators (APMs)

Related Skills