Voices of the Vigilant
Everybody wants AI right now, and that includes the teams security rarely sees as “technical.” When marketing, HR, and ops start downloading agent tools, prompt packs, and random code from the internet, we get a new kind of software supply chain risk, one that most security programs are not staffed or tooled to handle. I sit down with Amber Bennoui, a product leader and builder who has worked across cloud security, developer pipelines, and software supply chain security, and who now co-leads community efforts through the AI Security Alliance (AISECA). We talk about the mindset behind frontier work: learning fast, asking better questions, and refusing to ship “AI features” that do not answer the basics of who, what, when, where, and why. Amber shares what it looks like to pressure-test guidance with peer reviewers so it works in real companies, not just on a spreadsheet. We also go deep on Jiffy Labs, Amber’s project to bring visibility, scanning, and risk scoring to the AI artifact ecosystem. Think inventory for prompts, models, and agent components, plus practical ways to assess provenance and lineage when security tools are blind to what is actually being pulled into environments. From the Mythos conversation to the reality of ephemeral code rewritten by autonomous agents, we unpack why traditional security patterns struggle and why the AI “shared responsibility model” is still missing. If you care about AI security, AI governance, DevSecOps, and the future of AppSec, this conversation will sharpen how you think and what you ask for next. Subscribe, share the show with a friend, and leave a review to help more people find Voices of the Vigilant. Send us Fan Mail [https://www.buzzsprout.com/2434484/fan_mail/new] Support the show [https://www.buzzsprout.com/2434484/support] https://www.vigilantviolet.com/ www.linkedin.com/in/jessvachon1
21 episodios
Comentarios
0Sé la primera persona en comentar
¡Regístrate ahora y únete a la comunidad de Voices of the Vigilant!