#199 - How to Secure Generative AI

Join G. Mark Hardy in Torremolinos, Spain, for a deep dive into the security of Generative AI. This episode of CISO Tradecraft explores the basics of generative AI, including large language models like ChatGPT, and discusses the key risks and mitigation strategies for securing AI tools in the workplace. G. Mark provides real-world examples, insights into the industry's major players, and practical steps for CISOs to balance innovation with security. Discover how to protect sensitive data, manage AI-driven hallucinations, and ensure compliance through effective governance and ethical guidelines. Plus, get a glimpse into the future of AI vulnerabilities and solutions in the ever-evolving tech landscape.

References

OWASP Top 10 LLM Risks https://genai.owasp.org/

Gartner CARE Standard - https://www.gartner.com/en/documents/3980890

• Make sure your controls work consistently over time (Consistency)
• Make sure your controls meet the business needs (Adequacy)
• Make sure your controls are appropriate and fair (Reasonableness)
• Make sure your controls produce the desire outcome (Effectiveness)

Transcripts: https://docs.google.com/document/d/1V2ar7JBO503MN0RZcH7Q7VBkQUW9MYk6

Chapters

• 00:00 Introduction from Spain
• 00:42 Understanding Generative AI
• 03:25 Major Players in Generative AI
• 05:02 Risks of Generative AI
• 15:14 Mitigating Generative AI Risks
• 18:23 Implementing Solutions
• 24:09 Conclusion and Call to Action