AUTO-UPDATED

Are you sure that Kimmel clip was real? Late-night AI deepfakes are spreading online

AI-generated deepfakes of late-night talk show hosts like Jimmy Kimmel are proliferating on social media, exploiting viewer trust to spread potentially deceptive content and misinformation to online audiences.

Key Points

  • Generative AI tools allow creators to easily animate static images and clone voices of talk show hosts, who typically remain stationary on camera.
  • Experts note that these deepfakes gain credibility because they mimic the familiar context and style of legitimate late-night political commentary.
  • Platforms including YouTube, TikTok, and Meta require AI disclosure labels, yet many deceptive videos remain online for weeks, garnering hundreds of thousands of views.
  • Current legal protections are limited, as existing copyright laws do not specifically address the unauthorized creation of synthetic digital doubles of public figures.
  • Motivations for these videos range from ideological influence and political disinformation to generating ad revenue through clickbait.

Why it Matters

The rise of synthetic talk show content threatens to erode public trust in media by blurring the line between authentic political discourse and AI-generated disinformation. As these videos become more sophisticated and harder to detect, they complicate the efforts of social media platforms and regulators to protect the reputations of public figures and maintain an informed electorate.
NPR Published by Chloe Veltman
Read original