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!
--------
35:22
2025 LLM Apps Top 10, Evolving AI Architectures - GenAI Security Ep.1
Welcome to the first episode of the Gen AI Security Podcast! 🎉 Join Aubrey King, Sandy Dunn, and Scott Clinton as they discuss Thanksgiving plans and dive into the OWASP Top Ten for LLM apps. We'll cover our recent 2025 release, deepfake guidance, and evolving AI architectures like RAG. Perfect for a mix of holiday cheer and cutting-edge cybersecurity insights!
🔗 Resources:
OWASP GenAI Project: genai.owasp.org
#GenAISecurity #Cybersecurity #Podcast #Thanksgiving #AI #OWASP
Theme song, 'Crumbling Castle,' freely usable by King Gizzard And The Lizard Wizard, as per https://kinggizzardandthelizardwizard.com/bootlegger.
Polska