What Is a Deepfake?

Deepfakes are synthetic media, images, videos, or audio, generated by artificial intelligence to depict a real person saying or doing something they never actually said or did. The term combines "deep learning" (the AI technique used) with "fake." They are created using generative AI models trained on large collections of a target person's photographs or recordings.

How Deepfakes Are Created

The most common technique is a diffusion model or a GAN (Generative Adversarial Network). The model learns what a person's face looks like from dozens or hundreds of reference images, then generates new frames that map that face onto another body or create entirely new scenarios. In 2026, consumer-grade tools can produce convincing deepfakes in under an hour with fewer than 30 reference photos.

Why Non-Consensual Deepfakes Are a Federal Crime

The 2026 TAKE IT DOWN Act (47 U.S.C. § 223a) explicitly covers AI-generated intimate imagery. A synthetic nude or sexual video depicting your likeness is treated identically under federal law to an authentic photograph taken without consent. Platforms must remove such content within 48 hours of receiving a valid compliance notice.

How ScanErase Detects and Removes Deepfakes

ScanErase runs your facial biometric hash against 2.4 billion indexed face embeddings, including content that standard reverse image search misses entirely. When a deepfake is identified, your report gives you what you need to file a Section 223a compliance notice immediately. Platforms that fail to act within 48 hours can be escalated to the FTC.