Learn more about upcoming Events
From Multi-Agent Systems to Measurable Value
Experimental AI is over. In 2026, German industry moves from “What can agents do?” to “How do agents scale the bottom line?” Join our exclusive 5-session working group led by our Head of Agentic AI to explore the economics of Multi-Agent Systems—process impact, cost shifts, and real ROI.
AI's Blind Spots: From Prompt Injection to Rogue Agents
This Learn@Lunch is about security and reliability blind spots in modern AI. We cover prompt injection in LLMs, embedding leakage (vec2text), and risky agent behavior in real-world use. You’ll see concrete attack/failure examples and learn what to watch for when building or using LLM apps and agents.
Securing Agentic AI – From Threat Landscape to Adoption
AI reshapes cyber risk: agents speed attacks and fixes. Masterclass on securing agent deployments—protect data/products, prevent leakage, set guardrails (sandboxing, human-in-loop), secure auth + agent interop, and monitor end-to-end. Also covers cost control, compliant low/no-code, and appliedAI Working Group (Q3).
Countdown to the Transparency Deadline (Article 50)
EU AI Act transparency duties apply from 2.8.2026—most orgs aren’t ready. This 2‑hour masterclass answers: when must you disclose AI involvement? Mid‑2026 update, then Article 50: AI interaction notice, deepfake/AI labels, machine‑readable marking. Provider vs deployer clarified with examples, plus a roadmap checklist.
The agentic paradigm shift: Rethinking your AI Strategy
AI capability jumps every ~4 months; most orgs lag. Strategies for “AI as a tool” miss agentic AI, which runs end‑to‑end processes like a user. Add‑ons yield little; redesign around agents can slash cycle times. Hands‑on masterclass: a 3‑shift framework.