2-Step Agent: A Framework for the Interaction of a Decision Maker with AI Decision Support

We propose the 2-Step Agent framework, designed to optimize the interaction between human decision-makers and AI decision support systems. This framework structures the decision process into two distinct steps, enabling the AI to provide targeted, interpretable assistance while preserving human autonomy and cognitive load, leading to improved decision quality in various applications. This is highly relevant for real-world scenarios in business, medicine, and engineering where human experts collaborate with AI.