Continuous Red Teaming for AI: Insights from OWASP Experts - GenAI Security Ep.6
In this episode of the OWASP GenAI Security Podcast, Aubrey sits down with Ante and Kris from Splx to discuss the importance of continuous red teaming for AI applications. As AI evolves with multimodal systems, agent frameworks, and retrieval-augmented generation (RAG), proactive security testing becomes essential to address vulnerabilities like jailbreaks, data poisoning, and alignment shifts.Learn how integrating automated red teaming processes and collaborating closely with security teams can help organizations stay ahead of threats. Don’t miss this deep dive into the latest strategies for safeguarding AI applications in production and beyond.
--------
26:51
Exploring Agentic AI Security with OWASP: Insecure Agent Hackathon Apr 1 In NYC
In this episode of our podcast, Aubrey from the OWASP GenAI Security Project sits down with Ali Howe, the founder of Growth Cyber and contributor to the project. Alie shares her journey from software engineering to becoming a vCISO, emphasizing the importance of AI security and SOC 2 compliance for startups. She discusses the upcoming OWASP Insecure Agents Hackathon in New York City, where developers will explore building AI agents with a focus on identifying and mitigating security threats.Join us to learn more about the OWASP Gen AI Security Project, the significance of AI security in modern startups, and the exciting details of the hackathon event. Whether you're a seasoned developer or just starting out, this episode provides valuable insights and practical advice on enhancing your security posture with AI. Don't miss out on this engaging discussion!
--------
15:18
Sandboxing AI Models with Dyana & OWASP Top 10 for LLM Apps - Ep.4
Welcome back to our Generative AI Security Podcast! In this episode, Aubrey sits down with Ads Dawson, a respected member of the AI community, to delve into the OWASP Top Ten for LLM Applications and Generative AI. They discuss the latest security challenges and showcase Dyana, an open-source utility developed by Dreadnode for profiling and securing machine learning models.Ads gives a comprehensive demo of Diana, explaining its functionalities and how it can sandbox and profile a wide range of files, including models from Hugging Face. Discover how Diana can help detect backdoored models and provide detailed insights into GPU usage, network activity, and more. This episode is a must-watch for developers, security engineers, and anyone interested in the intersection of AI and cybersecurity. Don't forget to like, comment, and subscribe for more insightful discussions!
--------
22:36
Exploring AI Security: Red Teaming & Offensive Agents with Ads | GenAI Security Ep.3
Join us for the third episode of our podcast, where Aubrey from the OWASP Top Ten for Large Language Model Applications and Generative AI dives into the evolving world of AI security. In this episode, we chat with Ads, a seasoned AI security researcher from Dreadnode, who shares his journey from network engineering to AI security. Gain insights into the latest trends in AI red teaming, offensive agents, and the exciting developments at Dreadnode.
Explore the fascinating world of AI security with real-world examples, the importance of agentic AI, and how innovative tools like RoboPages and Burpference are shaping the future. Whether you're a security enthusiast or a professional in the field, this episode is packed with valuable information and engaging discussions. Don't miss out on the latest advancements and practical tips for navigating the Generative AI security landscape!
--------
32:19
Effective AI Red Teaming Best Practices With Krishna Sankar - GenAI Security Ep.2
In this second episode of our podcast series, we introduce the Red Teaming Initiative with special guest Krishna Sankar. Discover the nuances of red teaming in the context of generative AI, explore best practices, and understand what it takes to develop a mature red teaming strategy. Krishna also shares his rich background in autonomous systems, AI in financial services, and his passion for Legos!
Don't miss out on this informative discussion that also touches on the exciting (and somewhat daunting) concept of Artificial General Intelligence (AGI) and its implications.
For the latest insights and updates in generative AI security, make sure to like, subscribe, and stay tuned!
#GenerativeAI #LLM #RedTeaming #OWASP #ArtificialIntelligence #Podcast #AI #Security #AGI #AICommunity
Theme song, 'Crumbling Castle,' freely usable by King Gizzard And The Lizard Wizard, as per https://kinggizzardandthelizardwizard.com/bootlegger - PLEASE give them a listen!
Netherlands