Deepfakes, AI Influencers, and Why Verification Is Failing

Deepfakes, AI Influencers, and Why Verification Is Failing

From Dilemma Podcast by Jay Shapiro

April 30, 2026 · 1h 19m

About this episode

The episode explores the implications of deepfakes and AI-generated content on trust and authenticity in society.

How do we know what is real anymore?In this conversation with Professor Siwei Lyu of the University at Buffalo, one of the leading researchers in AI generated imagery and deepfake detection, we explore the unsettling world of synthetic images, cloned voices, fake influencers, and manipulated video. From AI girlfriends and scam artists to political propaganda and fake breaking news, we are entering a world where seeing is no longer believing.We talk about why “AI detectors” may be the wrong solution, why most people are already past the point of spotting fake images with the naked eye, and how verification itself is becoming unstable. If every image is already a processed representation of reality, what does authenticity even mean? We also discuss deepfake scams, revenge pornography, misinformation, metadata, provenance systems, and whether regulation can keep up with a technology that mutates faster than anyone can track.This is less about robot apocalypse fantasies and more about the quieter, more immediate danger, the erosion of trust, truth, and human connection. If AI is our generation’s splitting-the-atom moment, the real question is not whether the technology is good or…

People in this episode

Host: Jay Shapiro

Guest: Siwei Lyu

Topics covered

  • deepfakes
  • AI influencers
  • verification
  • synthetic images
  • misinformation
  • trust
  • authenticity

Keywords

  • deepfakes
  • AI influencers
  • verification
  • synthetic imagery
  • misinformation
  • trust
  • political propaganda

Mentioned in this episode

Organizations: University at Buffalo

More episodes of Dilemma Podcast

Explore listener stats, chart rankings, contacts and more on the Dilemma Podcast podcast page.