Paper Number

ECIS2026-2319

Paper Type

CRP

Abstract

Generative artificial intelligence (genAI) enables the next generation of information systems for strategic decision-making. Its abilities in retrieving, processing and generating knowledge could fundamentally reshape strategic intelligence, i.e., the acquisition and analysis of insights for strategic decision-making. Effective integration of genAI in strategic intelligence requires carefully designed human-AI collaboration to achieve complementary performance that exceeds the performance of either actor (i.e., human and genAI) alone. However, human-AI collaborative systems often still underperform compared to either actor working individually. It remains unclear what factors and mechanisms enable complementary performance in unstructured domains such as strategic intelligence. Therefore, we conducted a single-case study comprising five lab-in-the-field experiments and 25 interviews with analysts and managers. We develop a conceptual model comprising 13 factors that shape human-AI collaboration and propose six testable propositions that influence complementary performance. These results advance theoretical understanding of how complementary performance emerges while offering design-oriented knowledge for collaborative systems.

Share

COinS
 
Jun 14th, 12:00 AM

Designing For Complementary Performance: A Conceptual Model Of Human-AI Collaboration In Strategic Intelligence

Generative artificial intelligence (genAI) enables the next generation of information systems for strategic decision-making. Its abilities in retrieving, processing and generating knowledge could fundamentally reshape strategic intelligence, i.e., the acquisition and analysis of insights for strategic decision-making. Effective integration of genAI in strategic intelligence requires carefully designed human-AI collaboration to achieve complementary performance that exceeds the performance of either actor (i.e., human and genAI) alone. However, human-AI collaborative systems often still underperform compared to either actor working individually. It remains unclear what factors and mechanisms enable complementary performance in unstructured domains such as strategic intelligence. Therefore, we conducted a single-case study comprising five lab-in-the-field experiments and 25 interviews with analysts and managers. We develop a conceptual model comprising 13 factors that shape human-AI collaboration and propose six testable propositions that influence complementary performance. These results advance theoretical understanding of how complementary performance emerges while offering design-oriented knowledge for collaborative systems.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.