In the classic tale, Little Red Riding Hood is deceived by a wolf wearing a familiar disguise. In today’s enterprise environments, artificial intelligence can wear similar masks—sounding helpful, looking intelligent, but manipulated beneath the surface.

As AI adoption skyrockets, so do its vulnerabilities. Cyber adversaries are now exploiting this new attack surface, using adversarial prompts, data poisoning, and hallucinations to “dress up” threats as trusted systems. Just like Red was misled by a voice that seemed familiar, organizations are at risk of trusting models that can be subtly or even catastrophically compromised.

In this session, Jeff Crume will pull back that red hood on many AI attack types, demonstrate how threat actors are already exploiting them, and offer tactical ways to secure your AI systems.

Learning targets include:

• Recognize how AI systems can be deceived—just like Red Riding Hood—through adversarial and spoofing attacks.
• Explore six major forms of AI threats, including prompt injection, model inversion, and data poisoning.
• Learn how to “see through the disguise” and verify trust in AI tools before relying on them.
• Build organizational resilience by integrating AI-specific controls into your security strategy.