Securing Enterprise Agents — Private Dinner
Join Armorer Labs for an intimate, invite-only dinner bringing together CISOs, security leaders, AI executives, and founders building and securing the next generation of agentic systems.
An evening focused on a simple question:
How do we give AI agents enough autonomy to be useful while maintaining security, visibility, and control?
As agents gain access to internal data, APIs, codebases, tools, and critical systems, they introduce a fundamentally new security surface. The recent OpenAI agent incident involving Hugging Face offered an early glimpse into the kinds of risks that emerge when autonomous agents can interact with real-world systems.
We’ll bring together leaders across AI and cybersecurity for an off-the-record conversation about agent identity and permissions, tool access, prompt injection, data leakage, runtime security, observability, and governance—and what the security stack for the agentic enterprise needs to become.
A curated evening of great people, food, drinks, and candid conversation with those building and securing the next generation of AI systems.
Agenda
6 PM — Welcome and happy hour
6:30 PM — Dinner starts
Welcome and introductions around the table
Socratic dialogues: everyone in the room participates in the conversation
Topics
We’ll have Socratic dialogues where everyone has a chance to share what they’re seeing, challenge assumptions, and ask questions. In the spirit of AI Socratic events, the evening will be candid, technical, and laid-back.
We'll discuss:
The Agent Attack Surface — What changes when AI agents can call tools, execute code, access sensitive data, and take actions across production systems?
Identity, Permissions & Control — How should authentication, least-privilege access, delegation, and guardrails work when the user is an autonomous agent?
The Agent Security Stack — What does the security stack need to become to defend against prompt injection, data leakage, compromised tools, and malicious agent behavior at runtime?