In episode 185 of Cybersecurity Where You Are, Sean Atkinson sits down with Brian Calkin, Chief Technology and Innovation Officer at the Center for Internet Security® (CIS®); Theodore "TJ" Sayers, Senior Director of Threat Intelligence at CIS; and Kyle Leonard, Cyber Threat Intelligence Analyst at CIS. Together, they use a risk perspective to discuss artificial intelligence (AI) prompt injection and how to defend against it.

Here are some highlights from our episode:

• 00:49. A definition of AI prompt injection for businesses and executives
• 02:16. Brian on his role of guiding AI implementation at CIS
• 03:12. Understanding the urgency surrounding AI prompt injection as a security risk
• 05:32. Signals and trends indicative of threat actors attempting to weaponize prompt injection
• 07:10. How AI prompt injection differs from traditional input validation vulnerabilities
• 11:13. Early indicators that cyber threat intelligence (CTI) teams can monitor
• 15:00. The need to treat AI as a new identity in any enterprise implementation strategy
• 17:10. Understanding the difference: AI safety vs. AI security
• 20:36. Foundational, practical AI security that extends across all sectors
• 24:55. How CIS manages risk and supports the opportunity around the use of AI
• 28:25. The long-term promise of AI-driven vulnerability discovery grounded in fundamentals
• 34:48. Recommendations for piercing through the marketing hype surrounding AI

Resources

• Prompt Injections: The Inherent Threat to Generative AI
• New CIS Report Warns Prompt Injection Attacks Pose Growing Risk to Generative AI
• Episode 182: Striking a Balance on an AI Adoption Journey
• Episode 120: How Contextual Awareness Drives AI Governance
• Mythos AI: What Actually Matters for Cybersecurity Leaders
• Applying the CIS Controls to Real‑World AI Environments
• An Examination of Generative AI and Physical Threat Planning
• AI Playbooks for SLTT Cybersecurity Leaders

If you have some feedback or an idea for an upcoming episode of Cybersecurity Where You Are, let us know by emailing [email protected].