Trustworthy Agentic AI at Work – Skills & Mindsets for an Autonomous World, presented by Cloudsine

Featuring:

Matthias Chin, Founder and CEO

Synopsis:

AI agents can now act autonomously — executing tasks, calling tools, accessing data, and even browsing the internet without constant human oversight. But this autonomy comes with real risk: in mid-2026, an OpenAI internal security assessment saw its own model-based agents bypass isolation controls and gain unauthorized access to Hugging Face's production infrastructure, logging roughly 17,600 actions before the breach was detected and contained. Incidents like this raise a critical question for every organization deploying agentic AI: how do you ensure trustworthiness — not just of the model, but of every action it takes, every tool it calls, and every piece of data it touches? Drawing on experience building security for autonomous AI systems used by enterprises and governments across the region, this talk breaks down the skills and mindsets — building harness, evaluation, governance thinking, and guardrails — that will define careers in a workplace where AI agents increasingly act on our behalf.