Multi-Round Human-AI Collaboration with User-Specified Requirements

TL;DR

This paper introduces a principled, user-defined framework for multi-round human-AI collaboration that enforces constraints to improve decision quality while respecting human strengths and weaknesses.

Contribution

It formalizes counterfactual harm and complementarity principles via user-specified rules and presents an online algorithm with guarantees to enforce these constraints in collaborative settings.

Findings

The framework maintains harm and complementarity violation rates under dynamic interactions.

Adjusting constraints predictably influences human decision accuracy.

The approach is validated in medical diagnosis and pictorial reasoning tasks.

Abstract

As humans increasingly rely on multiround conversational AI for high stakes decisions, principled frameworks are needed to ensure such interactions reliably improve decision quality. We adopt a human centric view governed by two principles: counterfactual harm, ensuring the AI does not undermine human strengths, and complementarity, ensuring it adds value where the human is prone to err. We formalize these concepts via user defined rules, allowing users to specify exactly what harm and complementarity mean for their specific task. We then introduce an online, distribution free algorithm with finite sample guarantees that enforces the user-specified constraints over the collaboration dynamics. We evaluate our framework across two interactive settings: LLM simulated collaboration on a medical diagnostic task and a human crowdsourcing study on a pictorial reasoning task. We show that our online procedure maintains prescribed counterfactual harm and complementarity violation rates even under nonstationary interaction dynamics. Moreover, tightening or loosening these constraints produces predictable shifts in downstream human accuracy, confirming that the two principles serve as practical levers for steering multi-round collaboration toward better decision quality without the need to model or constrain human behavior.