Black Hat USA 2025: Reinventing Agentic AI Security with Architectural Controls @nccgroup
Black Hat USA 2025: Reinventing Agentic AI Security with Architectural Controls  @nccgroup
Uploaded September 2025 | Updated November 2026, 4 days ago
AI red teaming has demonstrated that guardrails alone cannot prevent prompt injection and may even create new attack vectors.

Drawing from extensive security testing of AI-driven applications, this talk highlights common vulnerabilities, showcases real-world exploits, and outlines architectural controls that neutralize injection-based risks. In this Black Hat 2025 briefing, you will learn practical design and threat modeling strategies to secure agentic AI systems against manipulation and data exfiltration.
Black Hat USA 2025: Reinventing Agentic AI Security with Architectural ControlsNCC Group Registry Data EscrowView - Portal FeaturesThe life of a Data Privacy ConsultantOT/IT Cyber Security Webinar Series - 3. Achieving 24/7 ResilienceThe Big Three webinar - Why is being proactive towards cyber resilience so important?NCC Groups Paul Vlissidis - Staying safe onlineFull Year Results Highlights Video for the 12 months ended 31 May 2016Meet the sales graduates in our escrow teamOT/IT Cyber Security Webinar Series - 2. Maximize OperationsCryptopals Guided Tour - 06 - Break repeating-key XORThe Big Three: Is there a cyber debt left by Covid-19? | Webinar recording | 23rd March 2021
NCC Group Global |

Black Hat USA 2025: Reinventing Agentic AI Security with Architectural Controls

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER