SEC.RHEIN by CLOUDYRION: AI Security Edition

SEC.RHEIN is a new after-work series by CLOUDYRION. We are a Düsseldorf-based cybersecurity consultancy specialising in Secure by Design and Ethical Hacking. The format is built around one idea: the best conversations in security happen outside conference rooms.

Each edition brings security professionals across disciplines into one room for short, sharp impulse talks followed by time to ask questions, discuss security and share experiences.

AI is reshaping both the threat landscape and the tools we use to defend against it. Attackers are leveraging AI to move faster and probe deeper. Defenders are exploring how the same capabilities can work in their favour. And the systems built on AI are introducing attack surfaces that the security community is only beginning to map. This edition puts AI Security in focus – from autonomous agents operating in industrial networks to the security implications of agentic AI entering everyday use.

Doors open at 17:30. Two impulse talks with moderated Q&A, a break in between, and plenty of time after the programme wraps to keep the conversation going over drinks and finger food.

This event is for cybersecurity professionals who want substance over slides and peer exchange over pitches. If you work in or around security and want to hear real takes on AI Security from people with hands-on experience, this is the room.

• Two impulse talks: 15–20 min + 5–10 min Q&A each
• Generous networking time with drinks and finger food
• Language: talks in English, networking in any language

Info: Industrial and enterprise networks accumulate assets over time – many of them unknown to the operators responsible for them. Unpatched and off the radar, these assets are exactly what attackers find and exploit as they move laterally through a network. Discovering them first is not optional.

In his talk, Mete walks through building an LLM-powered reconnaissance agent that discovers assets in an industrial network by combining traditional scanning with protocol-aware probing and web interface exploration. He covers the agent loop, tool-calling, prompt engineering for structured network reasoning, and what this reveals about where LLMs genuinely help – and where they still struggle.

Info: Agentic AI is no longer a research concept — it is already in consumers' hands. OpenClaw, a personal AI assistant designed to autonomously handle tasks from sending emails to booking flights, illustrates both the promise and the problem: systems with broad permissions, persistent access, and minimal friction are becoming mainstream before the security implications are fully understood.

In their talk, Nelli and Artem examine the security challenges that come with agentic AI adoption at scale. From data privacy over privilege escalation, to the attack surface that emerges when AI acts on behalf of users. They draw practical lessons from OpenClaw's design and the discussions it has already prompted, and translate them into concrete guidance: what cybersecurity professionals need to understand and do as agentic AI moves from early adopters to the enterprise.

Know what’s Happening Next — before everyone else does.

Get App

Host or Publisher CLOUDYRION GmbH