As AI moves beyond experimentation into real-world deployments where decisions may carry life-altering consequences, the industry needs systems for mitigating risk without sacrificing impact. This session examines the IRC’s approach in creating a hybrid human-AI adaptive loop architecture where agentic systems intelligently distinguish between interactions AI can safely handle and those requiring immediate human intervention, dynamically calibrating the level of human involvement to the stakes of each interaction. Participants will explore how this framework, built under the pressure of real constraints of humanitarian crises, offers a practical blueprint for responsible AI deployment in other high-stakes use cases.
Check out the incredible speaker line-up to see who will be joining André.
Download The Latest Agenda