Cybersecurity Awesomeness Podcast – Episode 171: AI Watermarking

In this episode of the Cybersecurity Awesomeness Podcast, host Chris Steffen, VP of Research at EMA, digs into Anthropic’s move to watermark AI-generated text and what it means for security professionals. Chris argues that while watermarking isn’t a bad idea, it’s an easily circumvented single point of verification, not a substitute for the “never trust, always verify” principle at the heart of zero trust. Drawing on real-world examples, from a hallucinated AI-generated map of the Las Vegas Strip to deepfake celebrity images and AI-fabricated legal citations, Chris makes the case that responsible AI use shouldn’t be penalized just because bad actors will always find workarounds. Tune in for a candid take on where watermarking fits — and where it falls short — in the broader conversation about content authenticity.