Science Friday Deepfakes Are Everywhere. What Can We Do?
Jan 22, 2026
Join experts Hany Farid, a UC Berkeley professor specializing in digital forensics, and Sam Cole, a journalist focused on deepfakes, as they navigate the murky waters of AI-generated media. They discuss the alarming ease of creating realistic deepfakes and their prevalence on social media, including the troubling rise of non-consensual images. Legal responses are fragmented, and they stress the need for regulatory action. Personal stories reveal the real harm deepfakes cause, including job loss and privacy invasion. Prepare to rethink what's real online!
AI Snips
Chapters
Transcript
Episode notes
AI Media Are Becoming Indistinguishable
- People are essentially at chance distinguishing real images from AI-generated images, and audio and video are closing fast.
- Hany Farid warns that visual content will soon be indistinguishable from reality and worsen misinformation dynamics.
Content Designed To Push Buttons
- Deepfakes are effective because they exploit a polarized and emotionally reactive online audience.
- Hany Farid notes that outrage-driven design amplifies confusion and lowers critical judgment while doomscrolling.
From Niche Tools To Phone Apps
- Sam Cole describes how deepfake creation moved from specialized computing to phone apps that need only one photo.
- She highlights how easy-to-use tools let anyone generate realistic fake images from a single social-media picture.