Two days of leadership talks and practitioner sessions focused on real-world AI security. Catch Ian Webster on the panel "Using AI for Offensive Security Testing."
Speaker Spotlight
Ian Webster
CEO & Co-founder, Promptfoo
Ian joined industry leaders for the panel "Using AI for Offensive Security Testing," covering how teams can use automation to discover LLM and agent vulnerabilities earlier in the lifecycle.
Key Themes
Critical topics shaping the AI security landscape in 2025 and beyond.
Adversarial AI
Understanding how attackers exploit LLMs through prompt injection, jailbreaking, and novel attack vectors targeting foundation models.
Defense Strategies
Building robust guardrails and implementing comprehensive red teaming programs to secure AI applications at scale.
Enterprise Readiness
Navigating compliance requirements, governance frameworks, and security best practices for production AI systems.
Future Threats
Anticipating emerging vulnerabilities in multimodal models, agents, and next-generation AI architectures.
Research Highlights
Automated Red Teaming
Demonstrated how open-source tools can systematically discover vulnerabilities in LLM applications through automated adversarial testing.
Try it yourself →Jailbreak Patterns
Analyzed common jailbreak techniques and their effectiveness across different model providers, revealing gaps in current safety measures.
View strategies →Data Exfiltration
Showcased novel methods attackers use to extract sensitive information from RAG systems and enterprise chatbots.
Explore plugins →Secure Your AI
Start red teaming your LLM applications today with Promptfoo's open-source security testing framework.