Agents Gone Rogue

Description

Agents Gone Rogue is an invite-only AI security evening for engineering leaders, taking place in Berlin on 25 August 2026, with talks, a panel and networking. As AI agents begin calling APIs, querying databases and interacting with production systems, organisations face urgent questions about access, oversight and cost. Hosted by [NetBird](https://netbird.io/?utm_source=luma), this focused event explores how technical leaders can keep agentic infrastructure under control,particularly when an agent has broader privileges than an employee or generates spending that nobody can easily attribute. The event runs from **18:00 to 22:00** at [Mindspace Krausenstraße](https://www.google.com/maps/search/?api=1&query=Mindspace%20Krausenstra%C3%9Fe&query_place_id=ChIJZXO14dBRqEcR0zJ2QRco9SM) in [Berlin](https://www.createwith.com/cities/berlin). It combines a practical technical talk, a candid panel discussion, audience Q&A and an extended opportunity to meet peers over food and drinks. ## Programme - **18:00 , Doors, drinks and networking** - **18:30 , Welcome** - **18:35 , Tech talk: “Kill switch for AI agents”** Jan Zboril, Site Reliability Engineer at Red Hat, and Martin Moucka, Senior Manager, IT Automation & Intelligence Evolution Team at Red Hat, will examine identity and access control for agentic infrastructure. The session covers how to control what an agent can reach, understand what it spends and cut off its access within seconds when something goes wrong. - **19:00 , Panel: “Taming AI sprawl”** A direct discussion about identifying which AI systems are actually running across an organisation, what they can access and what they cost, followed by audience Q&A. - **19:45–22:00 , Food, drinks and conversations** The panel brings together Martin Moucka of Red Hat; Ítalo Vietro, Senior Director of Engineering at Parloa; and Misha Bragin, CEO and Co-Founder of NetBird. Nima Sadeghifard, Chief of Staff at NetBird, will moderate. ## Who should attend? This evening is designed for people responsible for deploying and governing AI at scale: platform, infrastructure, DevOps and security engineers, alongside team leads and engineering managers. It will be especially relevant to those shaping company-wide AI adoption, managing production access or answering difficult questions when the AI bill suddenly doubles.… [Register for this event](https://lu.ma/ywjxiy8b)