Marketplace Tech Unraveling the complex knot of an AI-generated hoax
6 snips
Jan 22, 2026 
Casey Newton, a technology reporter and founder of the Platformer newsletter, dives into the world of AI-generated misinformation. He recounts a viral Reddit post that led him on a wild investigation, revealing how a source faked employee badges using AI. Casey highlights the importance of skepticism in the face of sensational claims and discusses the challenges reporters face in verifying authenticity amid increasing AI usage. With practical advice on maintaining journalistic integrity, this conversation is a must-listen for anyone navigating today’s digital landscape.
AI Snips
Chapters
Transcript
Episode notes
Fake Badge And 18‑Page Hoax
- Casey Newton received an 18-page document and a doctored employee badge from a purported whistleblower that seemed to corroborate viral Reddit claims.
- He later discovered the badge likely contained an AI watermark pointing to Google Gemini, revealing the materials were fabricated.
Check Images For AI Watermarks
- Use AI detection tools to inspect suspicious images because some generators embed identifiable watermarks.
- Check images with multiple models; Casey used ChatGPT and Google Gemini to find the Synth ID watermark.
AI Lowers The Cost Of Convincing Hoaxes
- AI has drastically lowered the effort needed to produce detailed, believable fabrications that once took significant time.
- Journalists and consumers must update verification practices because generating convincing hoaxes is now easy.