GenAI-assisted visual decision-making: a framework for human–AI collaboration

doi:10.1080/12460125.2026.2655904

GenAI-assisted visual decision-making: a framework for human–AI collaboration

2026-04-10

0

PRE

AI

DOI：10.1080/12460125.2026.2655904

Original

Original request for help

Save

Abstract

En 中文

Charts play an increasingly important role in decision-making across many fields. Although we know a great deal about how people process visual information, we know far less about how Generative Artificial Intelligence (GenAI) systems do so. Their multimodal models can now analyze charts and produce insights, yet they are often treated as autonomous decision-makers rather than as tools that complement human judgment. This paper takes a different perspective. We identify key strengths GenAI systems bring to visual analysis, including their ability to process large amounts of information quickly and to detect patterns humans may miss. Building on these strengths, we propose a human‑centered framework in which GenAI serves as a collaborative partner. The framework focuses on four tasks: aligning questions with data, aligning questions with visualizations, identifying design flaws, and validating, reconciling, and augmenting human insights. The quality of GenAI support for each task is evaluated through a series of experiments.

Keywords:

Generative Artificial Intelligence

human–AI collaboration

human-centered design

visual analysis

GenAI-assisted visual decision-making: a framework for human–AI collaboration

Abstract

Journal

Organization