A graduate methods course in detecting, verifying and tracing AI-generated text, images, video and audio. Every claim online is treated as a case that deserves a chain of custody, not a hunch.
The twelve weeks
Every lecture, every casework session, and every assignment deadline, in the order they actually happen. Each cell links to the real page.
What you will do
Every week you're handed a case: a passage, an image, a clip, a workplace document, a live dispute — and a growing toolkit for working it properly. You'll run textual and image forensics, embed and break watermarks, read a provenance standard instead of guessing, argue both sides of a real dispute before reaching a verdict, and finish the semester with an independent investigation you chose yourself. You walk out with a portfolio of worked cases and a habit of asking what would actually change your mind.
Who it is for
Graduate students who already write and read critically and want a concrete, evidence-based practice for the specific problem generative AI creates — not a general media-literacy refresher. You don't need a machine-learning background; you do need patience for a method that admits what it can't yet prove.