Knight Institute study finds minimal deepfake election impact

Knight First Amendment Institute documented only 78 deepfakes globally during 2024 elections, far below feared apocalyptic predictions about AI destroying democracy

Cheap fakes—edited videos without AI—appeared seven times more often than deepfakes according to News Literacy Project analysis of election misinformation

SKSteve Kramer received $6 million FCC fine for AI-generated Biden robocalls telling 🍁New Hampshire Democrats not to vote in January 2024 primary

Politicians increasingly claim authentic evidence against them is AI-generated, creating liar's dividend where real misconduct gets dismissed as fake

Social media algorithms failed to detect AI-generated political memes that received billions of views while reinforcing partisan beliefs rather than changing minds

Detection software promised by tech companies largely failed, leaving haphazard system of fact-checkers to identify synthetic content after damage was done

Foreign influence operations used AI through election day, spreading fake images of world leaders and voter fraud allegations in swing states