Zenity, the leading end-to-end security and governance platform for AI agents, has announced two upcoming AI Agent Security Summits taking place in London on October 8 and New York City on October 21. Organized and executed by Zenity Labs, the security research and community arm of Zenity, these events will bring together security researchers, practitioners and enterprise leaders in a vendor agnostic environment to examine emerging risks, attack techniques and practical approaches for securing autonomous AI systems.
Zenity announces AI Agent Security Summits in London and New York City.
The London summit takes place October 8 at 8 Bishopsgate.
The New York summit takes place October 21 at Pier Sixty.
The events are organized by Zenity Labs, Zenity's research and community arm.
The summits continue the global expansion of the series launched in 2025.
The London program covers agentic AI security policy and rogue agent behavior.
The summits continue the global expansion of Zenity's AI Agent Security Summit series, launched in 2025 to create a dedicated forum for the security community to share original research, real-world experiences and defensive strategies as AI agents become more deeply embedded across enterprise environments.
The London summit, taking place at 8 Bishopsgate, will explore the security implications of increasingly autonomous AI systems as organizations across Europe accelerate agent adoption. Sessions will address topics ranging from agentic AI security policy and rogue agent behavior to runtime risk assessment, multi-agent systems, adversarial attacks and emerging attack paths against AI-powered applications.
Featured London sessions and speakers include:
Daniel Cuthbert, Security Researcher and RUSI Associate Fellow, "The State of Agent Security: What, Why, When, and "Oh My God, They Didn't…"
Taylor Roberts, Jen Ellis, John Sotiropoulos and Kat Sommer, "Best Practice, No Backstop: Agentic AI's Public Policy Gap"
John Sotiropoulos, Founder and Principal Consultant, Deep Cyber, "Nine Seconds and No Attacker: Rogue Agency in the Enterprise"
Lorenzo Satta Chiris, Director of Excode, "AURA: Risk-Scoring Agent Actions Before They Execute"
Veli Oğuzcan Akdağ and Mustafa Bilgici, "Five-Stage LLM Agent Attack: From Passive Recon to Unauthenticated Prompt Exfiltration"
Yagmur Ceren Atay, "Agentic Failure Mode: How AI Defenders Become the Attacker's Best Assets"
Rico Komenda, "The Dark Side of Agentic AI: Adversarial Attacks at Scale"
Marius Hobbhahn, Co-founder and CEO, Apollo Research, "When Claude Goes Rogue"
Christian Schroeder de Witt, Associate Professor, University of Oxford, "Individually Secure Agents, Insecure Systems"
Michael Bargury, Co-founder and CTO, Zenity, "They Know They're Not Alone"
Two weeks later, the summit returns to New York City at Pier Sixty on October 21. Since the previous New York summit, the agent security landscape has changed significantly as organizations move beyond experimentation toward broader deployment of autonomous systems.
The New York program will examine three developments reshaping enterprise AI security: increasing agent autonomy, the democratization of agent development and the growing shared responsibility for AI security across security, development and business teams. As agents gain greater access to enterprise identities, data, tools and permissions, security leaders are increasingly being tasked with enabling AI adoption while establishing the visibility, governance and controls required to manage its risks.
Featured New York sessions and speakers include:
Helen Oakley, Leader at OWASP GenAI Security Project, "Who Has Authority in an Agentic Workflow?"
Sandeep Thuthike, Principal Engineer at Yahoo, "One Gate for Every Agent: A Control Plane for Human and Non-Human AI Identities"
Rahul Jain, Senior Director, Enterprise Applications AI & Technology at Gusto, "The Bypass Path: How Agent Security Controls Pass Every Test and Still Get Skipped"
Sumaiya Shrabony, Technical Program Manager and BI practitioner at the University of Colorado Denver, "Nobody Inventoried My Agents. I Am the Employee Your Security Team Keeps Warning You About."
Yotam Perkal, Security Research Lead at Pluto Security, "Vicious Circle: From a Filename to CI Pipeline RCE"
Yossi Eliaz, Principal Engineer and Head of DevRel at Incredibuild.com, "Your Agent Escaped Without Escaping the Sandbox"
Vikas Malik, Senior Lead Cybersecurity Architect at JPMorganChase, "The Agent Had Permission"
Ariel Fogel, AI Security Researcher, Office of the CTO at Pillar Security, "One Agent Called Another: Privilege Escalation in Google's ADK CI"
Tomer Elias, Product Executive and OWASP Workstream Leader, "Threat List to Build Plan: What's New in OWASP's Securing Agentic Applications Guide"
"The AI Agent threat landscape is vastly different today than it was over a year ago, when the first AI Agent Security Summit took place in NYC," said Michael Bargury, co-founder and CTO of Zenity. "Today, agents pose a threat to organizations through their connectivity, autonomy, and ability to execute, the same traits that make them so helpful to adopt. The Summit provides a platform for researchers to share their grounded testing, defenders to share their battle tested defense tactics, and leaders to share their vision of where this is going to continue to develop in the near future."
AI Agent Security Summit London
Date: October 8, 2026
Location: 8 Bishopsgate, London, UK
Information and registration: https://zenity.io/resources/events/ai-agent-sec...
AI Agent Security Summit New York City
Date: October 21, 2026
Location: Pier Sixty, New York City
Information and registration: https://zenity.io/resources/events/ai-agent-sec...
About Zenity
Zenity is the first security and governance platform purpose built for AI agents spanning SaaS, home grown platforms (Cloud) and end user devices (Endpoint). Trusted by Fortune 500 enterprises, Zenity helps security teams confidently adopt AI by delivering defense in depth with full lifecycle coverage, from agent discovery and posture management to real time detection, inline prevention and response. With an agent centric approach that prioritizes how agents behave, what they access and which tools they invoke, Zenity eliminates blind spots and enforces consistent policy and controls across environments so organizations can innovate with AI without compromising security.
