How Microsoft Uses AI for Threat Intelligence & Malware Analysis
What if the prompts used in your AI systems were treated as a new class of threat indicator? In this episode, Thomas Roccia, Senior Security Researcher at Microsoft, introduces the concept of the IOPC (Indicator of Prompt Compromise), sharing that "when there is a threat actors using a GenAI model for malicious activities, then the prompt... is considered as an IOPC".The conversation dives deep into the practical application of AI in threat intelligence. Thomas shares details from his open-source projects, including NOVA, a tool for detecting adversarial prompts, and an AI agent he built to track the complex money laundering scheme from a $1.4 billion crypto hack . We also explore how AI is dramatically lowering the barrier to entry for complex tasks like reverse engineering, turning a once-niche skill into something accessible to a broader range of security professionals .Questions asked:(00:00) Introduction(02:20) Who is Thomas Roccia?(03:20) Using AI for Reverse Engineering & Malware Analysis(04:30) Building an AI Agent to Track Crypto Money Laundering(11:30) What is an IOPC (Indicator of Prompt Compromise)?(14:40) MITRE ATLAS: A TTP Framework for LLMs(18:20) NOVA: An Open-Source Tool for Detecting Malicious Prompts(23:15) Using RAG for Threat Intelligence on Data Leaks(31:00) Proximity: A New Scanner for Malicious MCP Servers(34:30) Why Good Ideas are Now More Valuable Than Execution(35:30) Real-World AI Threats: Stolen API Keys & Smart Malware(40:15) The Challenge of Building Reliable Multi-Agent Systems(48:20) How AI is Lowering the Barrier for Reverse Engineering(50:30) "Vibe Investigating": Assisting the SOC with AI(54:15) Caleb's Personal AI Agent for Document OrganizationResources discussed during the call:NOVA- The Prompt Pattern MatchingDEF CON 33 Talk - Where’s My Crypto, Dude? The Ultimate Guide to Crypto Money Laundering
-------- Â
1:02:02
--------
1:02:02
The Future of AI Security is Scaffolding, Agents & The Browser
Welcome to the 2025 State of AI Security. This year, the conversation has moved beyond simple prompt injection to a far more complex threat: attacking the entire ecosystem surrounding the LLM. In this deep-dive discussion, offensive security experts Jason Haddix (Arcanum Information Security) and Daniel Miessler (Unsupervised Learning) break down the real-world attack vectors they're seeing in the wild.The conversation explores why prompt injection remains an unsolved problem and how the LLM is now being used as a delivery system to attack internal developers and connected applications. We also tackle the critical challenge of incident response, questioning how you can detect or investigate a malicious prompt when privacy regulations in some regions prevent logging and observability.This episode is a must-listen for anyone looking to understand the true offensive and defensive landscape of AI security, from the DARPA Cyber Challenge to the race for AI to control the browser.Questions asked:(00:00) Introduction(02:22) Who are Jason Haddix & Daniel Miessler?(03:40) The State of AI Security in 2025(06:20) It's All About the "Scaffolding", Not Just the Model(08:30) Why Prompt Injection is a Fundamental, Unsolved Problem(10:45) "Attacking the Ecosystem": Using the LLM as a Delivery System(12:45) The New Enterprise Protocol: Prompts in English(15:10) The Incident Response Dilemma: How Do You Detect Malicious Prompts?(16:50) The Challenge of Logging: When Privacy Laws Block Observability(21:30) Has Data Poisoning Become a Major Threat?(27:20) How Far Can Autonomous AI Go in Hacking Today?(28:30) An Inside Look at the DARPA AI Cyber Challenge (AIxCC)(40:45) Are Attackers Actually Using AI in the Wild?(47:30) The Evolution of the "Script Kitty" in the Age of AI(51:00) Would AGI Solve Security? The Problem of Politics & Context(59:15) Context is King: Why Prompt Engineering is a Critical Skill(01:03:30) What are the Best LLMs for Security & Productivity?(01:05:40) The Next Frontier: Why AI is Racing to Own the Browser(01:20:20) Does Using AI to Write Content Erode Trust?
-------- Â
1:24:46
--------
1:24:46
A CISO's Blueprint for AI Security (From ML to GenAI)
Is the current AI hype cycle different from the ones that failed before? How do you build a security program for technology that can't give the same answer twice? This episode features a deep-dive conversation with Damian Hasse, CISO of Moveworks and a security veteran from Amazon's Alexa team, VMware, and Microsoft.Damian provides a practical blueprint for securing both traditional Machine Learning (ML) and modern Generative AI (GenAI). We discuss the common pitfalls of newly formed AI Councils, where members may lack the necessary ML background to make informed decisions. He shares his framework for assessing AI risk by focusing on the specific use case, the data involved, and building a multi-layered defense against threats like prompt injection and data leakage.This is an essential guide for any security leader or practitioner tasked with navigating the complexities of AI security, from protecting intellectual property in AI-assisted coding to implementing safeguards for enterprise chatbots.Questions asked:(00:00) Introduction(02:31) Who is Damian Hasse? CISO at Moveworks(04:00) AI Security: The Difference Between the Pre-GPT and Post-GPT Eras(06:00) The Problem with New AI Councils Lacking ML Expertise(07:50) A History of AI: The Hype Cycles and Winters Since the 1950s(16:20) Is This AI Hype Cycle Different? The Power of Accessibility(20:25) Securing AI-Assisted Coding: IP Risks, Data Leakage, and Poisoned Models(23:30) The Threat of Indirect Prompt Injection in Open Source Packages(26:20) Are You Asking Your AI the Right Questions? The Power of "What Am I Missing?"(40:20) A CISO's Framework for Securing New AI Features(44:30) Building Practical Safeguards for Enterprise Chatbots(47:25) The Biggest Challenge in Real-Time AI Security: Performance(50:00) Why Access Control in AI is a Deterministic ProblemResources spoken about during the interviewTracing the thoughts of a large language model
-------- Â
52:16
--------
52:16
Gen AI Threat Modeling vs. AI-Powered Defense:
Is generative AI a security team's greatest new weapon or its biggest new vulnerability? This episode dives headfirst into the debate with two leading experts on opposite sides of the AI dragon. We 1st published this episode on Cloud Security Podcast and because of the feedback we received from those diving into all things AI Security, we wanted to bring it to those who haven't probably had the chance to hear it yet on this podcast. On one side, discover how to leverage and "tame" AI for your defense. Jackie Bow explains how Anthropic uses its own powerful LLM, Claude, to revolutionize threat detection and response. Learn how AI can be used to:Build investigation and triage tools with incredible speed. Break free from the "black box" of traditional security tools, offering more visibility and control. Creatively "hallucinate" within set boundaries to uncover investigative paths a human might miss. Lower the barrier to entry for security professionals, enabling them to build prototypes and tools without deep coding expertise. On the other side, Kane Narraway provides a masterclass in threat modeling the new landscape of AI systems. He argues that while AI introduces new challenges, many are amplifications of existing SaaS risks. This conversation covers the critical aspects of securing AI, including:Why access, integrations, and authorization are the biggest risk factors in enterprise AI. How to approach threat modeling for both in-house and third-party AI tools. The security challenges of emerging standards like MCP (Meta-Controller Protocol) and the importance of securing the data AI tools can access. The critical need for security teams to adopt AI to keep pace with modern engineering departments. Questions asked:(00:00) Intro: Slaying or Training the AI Dragon at BSidesSF?(02:22) Meet Jackie Bow (Anthropic): Training AI for Security Defense(02:51) Meet Kane Narraway (Canva): Securing AI Systems & Facing Risks(03:49) Was Traditional Security Ops "Hot Garbage"? Setting the Scene(05:57) The Real Risks: What AI Brings to Your Organisation(06:53) AI in Action: Leveraging AI for Threat Detection & Response(07:46) AI Hallucinations: Bug, Feature, or Security Blind Spot?(08:55) Threat Modeling AI: The Core Challenges & Learnings(12:26) Getting Started: Practical AI Threat Detection First Steps(16:42) AI & Cloud: Integrating AI into Your Existing Environments(25:21) AI vs. Traditional: Is Threat Modeling Different Now?(28:34) Your First Step: Where to Begin with AI Threat Modeling?(31:59) Fun Questions & Final Thoughts on the Future of AI SecurityResourcesBSidesSF 2025 - AI's Bitter Lesson for SOCs: Let Machines Be MachinesBSidesSF 2025 - One Search To Rule Them All: Threat Modelling AI SearchÂ
-------- Â
36:02
--------
36:02
Vibe Coding for CISOs: Managing Risk & Opportunity in AI Development
What happens when your product, sales, and marketing teams can build and deploy their own applications in a matter of hours? This is the new reality of "Vibe Coding," and for CISOs, it represents both a massive opportunity for innovation and a significant governance challenge.In this episode, join Ashish Rajan and Caleb Sima as they move beyond the hype to provide a strategic playbook for security leaders navigating the world of AI-assisted development. Learn how Vibe Coding empowers non-engineers to solve business problems and how you can leverage it to rapidly prototype security solutions yourself. Get strategies to handle the inevitable influx of AI-generated applications from across the business without overwhelming your engineering and security teams.Understanding the Core OpportunityAssessing the Real-World OutputManaging the "Shadow Prototype" RiskBuilding Proactive GuardrailsArchitecting for SafetyFor more episodes like this go to www.aisecuritypodcast.comQuestions asked:(00:00) Why Vibe Coding is a C-Suite Issue(02:34) The Strategic Advantage of Hands-On AI(04:20) Your AI Development Toolkit: Where to Start(12:08 Choosing Your First Project: A Framework for Success(16:46) The CISO as an AI Engineering Manager: A Step-by-Step Workflow(31:32) A Surprising Security Finding: AI and Least Privilege(36:47) Augmenting AI with Agents and Live Data(38:50) Beyond Code: AI Agents for Business Automation (Zapier, etc.)(43:30) The "Production Ready" Problem: Who Owns the Code?(53:25) A CISO's Playbook for Governing AI DevelopmentResources spoken about during the episode:AI Native Landscape - ToolsClineRoo-CodeVisual Studio CodeWindsurfBolt.newAiderv0 - VercelLovableClaude CodeChatGPT
The #1 source for AI Security insights for CISOs and cybersecurity leaders.
Hosted by two former CISOs, the AI Security Podcast provides expert, no-fluff discussions on the security of AI systems and the use of AI in Cybersecurity. Whether you're a CISO, security architect, engineer, or cyber leader, you'll find practical strategies, emerging risk analysis, and real-world implementations without the marketing noise.
These conversations are helping cybersecurity leaders make informed decisions and lead with confidence in the age of AI.
UK