About This Session
AI is evolving from static prompts to active agents: systems that don't just respond, but take action. In high-stakes sectors like healthcare and higher education, where there is no tolerance for error, the key question becomes: how can we implement this technology safely and responsibly? And what is already possible today?
What We Covered
A session organized by Shield vzw and hosted at C-mine Crib for a healthcare audience, diving deeper into the world of Agentic AI. We explored the latest trends, real-world applications, and the challenges that come with moving from AI experimentation to reliable implementation in clinical and care settings.
Key topics included:
- From prompts to agents — How the shift from reactive LLM responses to autonomous agent actions changes the risk profile for organizations
- High-stakes deployment — What safe implementation looks like in healthcare, where errors carry real consequences for patient safety and data privacy
- What's possible today — Practical examples of agentic AI already in production, and where the boundaries currently sit
- Responsible implementation — Governance frameworks, guardrails, and the human oversight patterns that make agentic deployment viable