Synthetic media incident archive
When fake video
did real damage.
A running record of synthetic media used to deceive: deepfake videos, cloned voices, and AI-generated images that moved markets, swung headlines, and stole millions. For each one: what happened, what was faked, and how it was caught.
This archive documents confirmed, widely-reported incidents. It is a record of the public signals around each case, not a forensic re-analysis. Dates reflect when each incident surfaced publicly.
Google's Veo 3 raises the floor on convincing fakes
When AI video crossed from obviously fake to genuinely hard to call.
Read the case →The $25 million deepfake video call
A finance worker wired a fortune after a meeting where everyone but him was fake.
Read the case →The fake Biden robocall before the New Hampshire primary
A cloned presidential voice told voters to stay home. It cost its maker millions.
Read the case →Explicit Taylor Swift deepfakes flood social media
Non-consensual AI images reached millions before platforms could react.
Read the case →Tom Hanks warns of an AI version of himself
A wave of synthetic celebrity endorsements sold products the stars never touched.
Read the case →A fake Pentagon explosion briefly rattles the market
One AI image of an event that never happened moved real money.
Read the case →The fake Zelensky "surrender" video
An early wartime deepfake tried to fake a national surrender, and failed.
Read the case →Tracking the record. This archive is curated from confirmed, widely-reported cases and is updated as significant incidents emerge. For broader, continuously-updated coverage, these trackers are useful starting points:
Deepfakes Tracker · Resemble AI Incident Report · Lead Stories Deepfakes · Authenticity Crisis
Spotted an incident we should add, or watching a clip you are unsure about? Run it through the analyzer or send it to us.