AI Security Lightning Talks – Heron Delegation @ LISA

​Heron, an AI Security field-building organization, is excited to bring a delegation of cybersecurity experts to London for a 3-day seminar hosted at London Initiative for Safe AI (LISA).

​Join us for a part of our programming open to security-curious public: four 10-minute lightning talks at the intersection of AI Safety x Cybersecurity!

​🔷 Eden Katz (Alice)
💪 RL Gym Blueprint: How To Align AI Against Threats
Eden is a GenAI Security Tech Lead at Alice, leading red-teaming and black-box operations for frontier models. She specialises in discovering novel abuse techniques within AI ecosystems, focusing on agentic workflows and nation-state threat vectors.

​🔷 Inbar Shulman (Independent)
❔What Threat Modeling Actually Is
Inbar is a lead R&D in cyber and algorithms, overseeing AI, data, and cyber strategy at the General Staff level. He holds an MSc in Computer Science, with specialization in formal methods, and is currently pivoting into AI safety full-time.

​🔷 Gal Wiernik (Founder)
🔍 Whose Side Is Your AI On? Secret Loyalties and AI Integrity
Gal is a technical founder with 7 years of hands-on national cybersecurity offense and defense research and engineering. He is building an AI Safety/Security startup focused on model security, integrity, and adversarial ML.

​🔷 Guy Nachshon (Founder)
⚒️ agent-JEPA: Training World Models for AI Control
Guy is a Heron Fellow researching eval awareness and a technical founder, interested in building world models for critical infrastructure. He previously co-founded an AI supply chain security startup, using mechanistic interpretability for deep model analysis.