Agentic AI Security Summit, Europe: OWASP Community Overview

Описание: This closing talk from the OWASP Agentic Security Summit (London, Dec 9, 2025) brings the focus back to community and why OWASP is critical to securing generative and agentic AI. Sam Stein (OWASP London Chapter Leader & OWASP Global Board) explains that while AI security builds on traditional AppSec foundations, non-determinism, natural language inputs, hallucinations, and autonomous behavior create entirely new risks with no guaranteed fixes.

Through a real-world chatbot jailbreak case study, Sam highlights a major industry gap: many organizations still dismiss prompt injection, hallucinations, and harmful model behavior as “non-security issues.” This legacy mindset, combined with unclear ownership, immature standards, limited tooling, and pressure to ship AI fast, leaves companies exposed—especially as systems move toward autonomous agents.

The key takeaway: this is where OWASP and the GenAI Security Project matter. OWASP provides open, community-driven frameworks, taxonomies, testing approaches, and guidance to make AI security measurable, actionable, and shared across teams. Join local chapters, contribute to projects, attend events, and help shape secure AI practices together.

🔗 Learn more: https://genai.owasp.org

Hashtags:
#OWASP #GenAISecurity #AgenticAI #AISecurity #AppSec #CyberSecurity #LLMSecurity #AIThreats #PromptInjection #OpenSourceSecurity #SecurityCommunity