What we’re about
Welcome to the AI Security Collective Meetup group! We are a community of professionals, researchers, and enthusiasts who are passionate about artificial intelligence and cybersecurity. Our group focuses on exploring the intersections of AI and security, discussing the latest trends and developments in the field, and sharing knowledge and best practices. Whether you are an expert in the field or just starting out, this group provides a platform for networking, learning, and collaborating on cutting-edge AI security research and technology. Join us to connect with like-minded individuals and stay at the forefront of the rapidly evolving AI security landscape.
Upcoming events (1)
See all- Security for AI AgentsControl Plane, London
Join us for the September AI Security Collective meet-up, where we will be diving into security of AI Agents!
This month we're joined by Luke Hinds who will sharing AgentUp, an open source project, being built to make Agents safe, portable, robust and secure.
Schedule:
- 1800: Doors Open
- 18:20: Kick Off Introduction
- 18:30: Zero Trust AI Agents: AgentUp's Scope-Based MCP and Tool Security with Luke Hinds (Founder, Red Dot Rocket and creator of SigStore)
- 19:15: Expert Panel: Securing AI Agents Mike Lieberman (CTO, Kusari) Luke Hinds (Founder, Red Dot Rocket) and Hanah-Marie Darley (Founder and Chief AI Officer)
- 20:00: Networking
- 21:00: Home Time
Speaker:
Luke Hinds is founder of Red Dot Rocket and creator of AgentUp, a developer-first framework that brings production-grade security and standardization to AI agents.He is the founder of Sigstore, the open source project that revolutionized software supply chain security by making cryptographic signing accessible to millions worldwide.
Previously a Distinguished Engineer at Red Hat, and the co-founder and CTO of Stacklok, inc. He has held numerous community leadership roles, including the Kubernetes Security Team and technical steering committee chair for Sigstore.
Talk: Zero Trust AI Agents: AgentUp's Scope-Based MCP and Tool Security
AI agents with unrestricted MCP and Tool access can be the source of huge security risks. Traditional permission models weren't designed for autonomous systems that can chain tool calls and access multiple data sources dynamically. Most frameworks built to date have been made by developers with very little knowledge of systems security. This talk demonstrates AgentUp's zero trust approach to AI agent security through fine-grained scope controls. You'll learn how scope-based permissions contain MCP server access, limit tool capabilities, and prevent privilege escalation. AgentUp is an open source project, being built to make Agents safe, portable , robust and secure: github.com/RedDotRocket/AgentUp
Panellists:
Michael Lieberman is co-founder and CTO of Kusari. His passion is in applying his expertise to use cases where privacy and security are paramount. Mostly recently he has been focused on work within the software supply chain security space. He is also highly committed to open-source, having co-created projects like GUAC and FRSCA, along with having co-lead white papers like the CNCF’s Secure Software Factory Reference Architecture.
Hanah-Marie Darly brings a uniquely human-centered perspective to AI governance. Her academic foundation in psychology and international relations informs the design systems that anticipate both technological and behavioural risks - ensuring alignment between AI actions and human intent. With over a decade of experience in threat intelligence, geopolitical analysis, and applied psychology, Hanah-Marie is how working on an early-stage venture helping enterprises adopt agentic AI in a secure, governed, and explainable way.
Thanks to our sponsors Mindgard and ControlPlane for supporting this community!
Registration Questions: Please note that the details you provide on registration will be shared with our sponsors who may wish to follow up with you after the event.
