CTI Session: Adversarial prompts. Open-source tools that detect, hunt, and block IoPCs
Threat Intelligence

CTI Session: Adversarial prompts. Open-source tools that detect, hunt, and block IoPCs

Wednesday, November 5th, 8:00AM GMT

Josh Darby MacLellan
on Oct 30, 2025

Generative AI is now embedded in business workflows, along with malicious prompts. Adversarial prompts, or Indicators of Prompt Compromise (IoPCs), reveal attempts to exploit, abuse, or manipulate AI models for malicious purposes. Join Thomas Roccia, threat researcher, author, and speaker, to explore how security teams can detect and defend against these emerging threats using open-source tools he’s created.

Register

You'll learn:

  • What makes adversarial prompts dangerous: How IoPCs differ from traditional IoCs and the techniques attackers use to exploit LLMs
  • How to detect malicious prompts with NOVA: A Python-based detection tool using keyword, semantic, and LLM-based matching to identify adversarial patterns
  • How to share intelligence with PromptIntel: A brand new community platform for submitting, verifying, and sharing real-world IoPCs with detection rules

Walk away with practical techniques to operationalize open-source tools and share intelligence, so you can detect and block adversarial prompts before they impact your systems.

Webinar - Adversarial prompts: Open-source tools that detect, hunt, and block IoPCs

Wednesday, November 5th, 8:00AM GMT

Register
Espace publicitaire · 300×250