Essays on artificial intelligence and creative problem solving

Guide(s)

De, Rahul; Banerjee, Shankhadeep

Department

Information Systems

Area

Information Systems

University

Indian Institute of Management Bangalore

Place

Bangalore

Publication Date

3-31-2026

Year Awarded

March 2026

Year Completed

March 2026

Year Registered

June 2020

Abstract

Creative problem-solving (CPS) is crucial for the growth of organizations and individuals. For a long time, it has been viewed as a purely human skill, but over the period, the field of creativity interacted with technology, which gave rise to a variety of creative systems such as Creativity Support Systems, Co-creative Systems, and Autonomous Creative Systems (Davis et al., 2015; Rezwana & Maher, 2023). In today’s era of sophisticated and highly agentic technologies, these systems have evolved from mere tools to the drivers of human action, thereby transforming traditional processes. The advent of Generative Artificial Intelligence (GenAI) technology and its quick permeation into our daily lives have triggered an interesting debate about how GenAI-based systems impact creative problem-solving. On the one hand, these systems offer tremendous potential to enable enhanced and distributed problem-solving through cognitive load distribution, content generation, information processing and reasoning, and human-like problem-solving (Boussioux et al., 2024; Krakowski, 2025; Orrù et al., 2023), whereas on the other hand, they are found to cause overreliance, disrupt workflows, reduce situational awareness, and hamper people’s critical thinking (Candelon et al., 2023; H.-P. H. Lee et al., 2025; Simkute et al., 2024). These contradictions signal a theoretical puzzle about the duality of AI’s creative potential. The competing mechanism involves AI’s augmentative capability to support human cognition versus its restrictive potential to anchor human thinking. While the strong capabilities of these systems are undeniable and therefore make them unavoidable, the potential adverse impacts highlight a need for the conscious and regulated use of these systems. This phenomenon-based research explores how GenAI systems are transforming various aspects of creative problem-solving phenomena. Creative problem-solving has wider implications and better generalizability as it has been applied successfully in various settings and by problem solvers of all ages (Treffinger, 1995). GenAI is becoming a ubiquitous technology moving towards artificial general intelligence (Bubeck et al., 2023; Krakowski, 2025). Therefore, this research exploring the interaction of GenAI and CPS is of high importance to both industry and academia. Essay 1 is a conceptual study. Here, we develop a human-IS CPS framework grounded in the dynamism of human-system interaction by deriving various interrelated constructs and assumptions involved in human-creative system collaboration from a meticulous synthesis of the philosophy of creativity and creativity with computational systems literature streams. By integrating diverse systems under a unified structure, the framework enables meaningful comparisons and reduces fragmentation in existing literature. The study uses Rhode’s 4P model to provide an underlying structure to the framework and the theoretical lens of ‘agency’ to differentiate the fundamental nature of creative systems. Then, by applying this framework to the context of GenAI-based creative systems, this study provides interesting insights into how various aspects of the CPS phenomenon are being influenced. The specific research questions answered in this essay are: 1) How do humans and creative systems collaborate for creative problem-solving? What are the key constructs and their relationships in holistically modelling this sociotechnical phenomenon? 2) How have GenAI-based creative systems changed the co-creative phenomenon (as modelled in research question 1)? The framework comprises the key constructs such as humans, creative systems, creative problems, creative process, outcome, external factors, human-system collaboration spectrum, evaluation-feedback loop, and various fits, e.g., human-problem fit, human-process fit, system-problem fit, and system-process fit. Further, the answer to RQ2 highlights a fundamental reconfiguration of roles, processes, and outcomes as well as the emergence of new external factors. Essay 2 is an empirical study where we examine the impact of collaborating with GenAI for CPS on the crucial dimensions, i.e., novelty and value, of the creative outcome. Further, it also examines the moderating effects of the problem solvers’ perceived expertise, creative self-confidence, and creative ability to understand how the overall effect varies among people with different levels of metacognition and abilities. Furthermore, the effects of GenAI are also benchmarked with a widely used search engine. We use online experiments based on Latin-Square design on the Prolific platform to answer the research questions: 1) What is the impact of human-GenAI collaboration on the outcome of creative problem-solving, i.e., novelty and value of creative ideas? 2) How does this impact vary between users of different levels of creative ability and creative confidence? The results demonstrate the detrimental effects of GenAI on the novelty of ideas, whereas it enhances ideas’ value. Further, the adverse impact on novelty prominently affects people with less expertise, less confidence, and high creative ability. Overall, GenAI reduces the gap in idea novelty between people with low and high confidence and creative ability. Interestingly, this gap reduction is resulting from a decline in high-performers’ outcomes rather than an improvement in the outcomes of low-performers. Essay 3 investigates how GenAI usage impacts problem solvers’ metacognitive interpretation of the creative problem, which can be a prospective mechanism through which creative systems impact the CPS outcome. The specific research questions are: 1) What is the impact of collaborating with GenAI and non-GenAI systems for CPS on problem-solvers’ perception of problem difficulty? 2) How does this impact differ between people with different levels of metacognitive judgment, i.e., expertise and task-based self-confidence? The results demonstrate that overall, both GenAI and non-GenAI systems alleviate problem solvers’ perception of difficulty; however, GenAI does it better than non-GenAI systems. Additionally, the use of GenAI reduces the gap in perceived problem difficulty between less confident and more confident people when compared to no technology use. The additional analysis also establishes the perception of problem difficulty as a prospective mediating mechanism through which GenAI-based creative systems impact the creative outcome. These findings suggest that integrating advanced AI tools can foster more inclusive and effective creative problem-solving environments. Theoretically, this research contributes to the mainstream IS research by developing a comprehensive human–IS framework for CPS; it introduces a new classification of creative systems based on agency and highlights the dynamic, context-dependent nature of human–AI collaboration, calling for extensions to existing theories like task-technology fit that consider well-defined problems. Additionally, it extends traditional idea generation models (CNM and SIAM) by showing how GenAI-specific effects like anchoring can interfere with creativity. It also questions the idea of democratization of creativity, cautioning that equalization may come at the cost of high performers' outcomes. Finally, it contributes to theories of cognition, self-efficacy, and flow by showing how GenAI affects users’ metacognition and supports co-creative experiences, offering new pathways for IS and HCI research. Practically, our comprehensive framework can help organizations to plan resources, align human–AI collaboration, and deploy GenAI where it adds the most value, i.e., in idea evaluation. The study advises designers to balance convergence and divergence in GenAI tools to better support creativity, and guides managers on customizing GenAI use per employees’ creative confidence and creative ability. It also highlights GenAI's potential to reduce mental barriers, which may help in fostering inclusivity in organizations. Lastly, the study stresses the importance of careful collaboration to avoid cognitive biases, and encourages organizations to invest in AI literacy and develop policies for responsible AI use aligned with global standards.

Pagination

xix, 247p.

Copyright

Indian Institute of Management Bangalore

Document Type

Dissertation

DAC Chairperson

De, Rahul; Banerjee, Shankhadeep

DAC Members

Mishra, Sushanta Kumar

Type of Degree

Ph.D.

Relation

DIS-IIMB-FPM-P26-07

This document is currently not available here.

Share

COinS