Enterprise AI is moving faster than teams can prove it is safe. Yash Jhunjhunwala shows how Qualys TotalAI discovers shadow AI, tests it, and protects AI agents from code to runtime. Join us at ROCon Americas, October 13-15 in Austin, Texas: https://www.qualys.com/rocon/2026/americas Try Qualys ETM free for 30 days: https://www.qualys.com/free-trial-new/enterprise-trurisk-management This Black Hat USA 2026 session covers the AI security lifecycle in three parts: discover AI workloads, software, agents, and MCP servers across cloud and on-prem; assess them with adversarial testing such as prompt injection and jailbreaks plus posture management for SaaS AI tools; then prioritize and remediate with business context and guardrails. It also ties AI risk to frameworks and regulations including the EU AI Act, the OWASP Top 10, and MITRE ATLAS. The demo scans AI code repositories down to the line of code, follows the same agent into containers and hosts, and then shows an AI agent running in production connecting to external AI APIs. A browser plugin also checks employee prompts before they reach AI tools. Speaker: Yash Jhunjhunwala, Lead SME, Cloud Security Solutions and AI Security, Qualys Chapters: 0:00 Intro: enterprise AI is outpacing security 1:27 Discover, assess, then protect 1:46 Industry trends: prompt injection and shadow AI 2:36 Regulations: EU AI Act, OWASP, and MITRE ATLAS 3:31 Qualys TotalAI 2.0 overview 4:45 Securing AI from code to runtime 5:32 Securing the workforce: browsers and SaaS AI 6:25 Prioritizing and remediating AI risk 7:33 Demo: code repositories and compliance findings 9:06 Demo: deployed workloads 9:47 Demo: AI agents in runtime and browser activity 11:02 Join us at ROCon Americas #BlackHat2026 #AISecurity #AgenticAI

 71      0