Genetec's Mathieu Chevalier on tricking AI and what Claude Mythos really means for the industry

Genetec's lead security architect joins the Cyber Uncut podcast to unpack the impact of AI on vulnerability hunting, prompt injection, and how to become invisible with one simple trick. This week on the Cyber Uncut podcast, host David Hollingworth speaks with Genetec's lead security architect, Mathieu Chevalier, about how AI is dramatically impacting both physical and cyber security, the hidden costs of Anthropic's Claude Mythos, and why AI speed must always be matched with human intelligence. But first, Chevalier shares his simple trick to appear invisible to AI with the assistance of just one page of text – or maybe just a T-shirt! The pair then dive into how AI empowers Genetec's security platforms, the rising costs of every AI prompt, and the true power of vulnerability hunting in a post-Mythos world. "I would say the best practices are not established now. It's still so new, and changing so fast," Chevalier says. "So we're kind of trying to learn all together, as a cyber security community, how to do this." Enjoy, The Cyber Uncut team