Case Study: Detecting Deepfakes and AI-Generated Media at Platform Scale
Design a trust-and-safety system that flags synthetic images and video across hundreds of millions of daily uploads — a defence-in-depth cascade of provenance, hashing, learned detectors and human review, and this block's mirror image of the GAN foundation subject
Model interview answer for platform-scale synthetic-media detection (the trust-and-safety systems behind Meta, YouTube, TikTok-class platforms, and the KYC-selfie-liveness vendors banks buy from): the block's adversarial case and the deliberate mirror of the GAN foundation subject — the very StyleGAN internals that subject teaches to generate are what a frequency-domain forensic detector here learns to spot. A defence-in-depth cascade argued as the thesis: C2PA content credentials and SynthID-class invisible watermarks prove an asset is known-real or known-generated cheaply and deterministically (provenance), which is a different problem from learned detectors proving an unlabelled asset is fake (detection), and a platform needs both. The DFDC's published generalization trap — leaderboard detectors that collapse on an unseen black-box generator — as the reason an in-house generator zoo and unseen-generator AUROC splits are the only honest way to build and grade a detector. Frequency-aware CNNs, face-crop pipelines, video frame-sampling budgets, multimodal lip-sync detectors, and calibration for a score rather than a verdict. A decision-policy section mirroring the fraud-detection case's review-capacity framing, and a scalability deep-dive built around the cascade-as-architecture pattern and continuous retraining as an arms race against new generator families.
Practice questions (6)
-
View →
Why In-Distribution AUROC Is the Wrong Headline Metric
Advanced · Free -
View →
Provenance and Detection Are Not the Same Layer
Advanced · Free -
View →
Reach-Weighted Review Prioritisation
Advanced -
View →
Frequency-Domain Detection and Why It Traces to StyleGAN's Architecture
Advanced -
View →
Video Frame-Sampling Budgets and What They Trade Away
Advanced