Back to AIBriefs
AnalysisPolicyCybersecurity
Featured·

Red-Teaming after Mythos: Gray Swan cofounders on AI security

Zico Kolter and Matt Fredrikson discuss AI red teaming and prompt injection, referencing their tool Shade used by Anthropic. They cover agent vulnerabilities, automated red teaming, and enterprise guardrails.

Jun 22, 9:06 PM