Перейти к содержимому

Free Class OCT 13: RED TEAM labs for attacking OWASP top 10 LLM

Dean Bushmiller

0:00 / 0:00

Free Class OCT 13: RED TEAM labs for attacking OWASP top 10 LLM

27 просмотров · 7 дней назад
Dean Bushmiller
357 подписчиков
27 просмотров · 7 дней назад
Sign up free October 13 https://maven.com/p/19c559/labs-for-a... Think your system prompts are enough to secure "vibe coding"? Think again. With everyone—and their mom—generating code with AI, internal programming isn't under control, and attackers aren't running out of time. In an environment where every security professional must be on the Red Team, traditional defenses are falling short. Join Dean Bushmiller (CISSP, SecAi+, 38+ years experience) for a free, hands-on 30-minute live masterclass focused on Attacking the OWASP Top 10 for LLMs. Rebuild your security playbook while you row the boat in real time. 🚨 FREE LIVE LAB MASTERCLASS: Attacking OWASP Top 10 for LLMs 📅 Date: Tuesday, October 13 ⏰ Time: 10:00 AM Central ⌛ Duration: 30 Minutes 👉 Reserve Your Seat: [INSERT SIGNUP LINK HERE] 🎯 WHO THIS CLASS IS FOR Cybersecurity Professionals & Red Teamers: Needing hands-on exposure to real-world LLM exploitation vectors. Non-Programmers & Tech Leaders: Who need to understand AI threat mechanics without needing to write code from scratch. Flexible Learning Styles: Experience the labs in 3 formats tailored to your style: Watching, No-Code, or Programmer. 💡 WHAT ATTACKS YOU WILL LEARN & DEMO 1. Poisoning What the AI Learns (Context & Knowledge Poisoning) How feeding false facts—without hidden commands—completely subverts AI reasoning and decision-making. 2. Hijacking What the AI Reads (Indirect Prompt Injection & Vision Attacks) Hiding prompt injections inside web pages that AIs process but humans can't see. Bypassing OCR and vision models with visual tricks. Real-World Impact: How a single hidden line of text gets a fraudulent invoice automatically approved. 3. Abusing What the AI Can Do (Agent & MCP Exploitation) Tricking autonomous AI agents into executing unauthorized actions and tool misuse. Hijacking Model Context Protocol (MCP) tool calls to penetrate internal architecture. 🧪 HANDS-ON LAB SETUP & ASSETS Zero-Cost Local Setup: Test your machine's capabilities and run our free lab environment directly on your device. 3 Track Options: Participate by Watching, running No-Code modules, or deep-diving as a Programmer. Visual Diagrams: Includes full architecture diagrams for every attack scenario covered.