This theoretical paper examines how intelligent knowledge management systems (IKMS) enable human–AI collaboration in financial decision-making and how such collaboration contributes to value creation beyond profit. The study addresses a theoretical gap at the intersection of behavioural finance, artificial intelligence in finance, knowledge management, and organisational innovation. Behavioural finance has demonstrated that financial judgement is shaped by biases, heuristics, emotions, and bounded rationality, while AI research in finance has largely emphasised automation, predictive accuracy, efficiency, and performance optimisation. However, less attention has been paid to how human cognition and algorithmic intelligence interact within organisational knowledge processes, and how this interaction reshapes financial decision-making in ways that affect trust, accountability, sustainability, and innovation. Using a conceptual framework-building approach, the paper positions IKMS as the central organisational layer through which human and AI cognition are integrated. Financial decision-making is conceptualised as a knowledge-intensive organisational process involving knowledge retrieval, knowledge integration, knowledge interpretation, and knowledge application. Within this process, AI contributes computational capacity, pattern recognition, prediction, and scalable knowledge synthesis, while human actors contribute contextual judgement, ethical reasoning, experience-based interpretation, and responsible oversight. The framework argues that human–AI collaboration becomes valuable not when AI replaces human judgement, but when IKMS enable complementarity, augmentation, validation, and continuous learning between human and machine intelligence. A central contribution of the paper is to show that AI does not eliminate behavioural bias in financial decision-making. Rather, biases may be redistributed, amplified, or transformed within hybrid human–AI knowledge systems. For example, AI-supported financial recommendations may reduce some forms of human overconfidence while creating new risks of automation bias, anchoring on algorithmic outputs, or excessive trust in model-generated classifications. The framework therefore identifies governance, explainability, trust calibration, ethical alignment, and feedback mechanisms as critical boundary conditions for responsible and effective human–AI collaboration. The paper conceptualises value creation beyond profit across four interrelated dimensions: cognitive value, reflected in improved decision quality, bias recalibration, and enhanced risk insight; institutional value, reflected in transparency, accountability, and stakeholder trust; societal value, reflected in ethical alignment, financial inclusion, sustainability, and long-term stability; and organisational innovation, reflected in improved knowledge recombination, adaptive learning routines, and more responsive financial decision processes. By integrating behavioural finance with intelligent knowledge management systems, the study contributes to a more comprehensive understanding of how AI-supported financial institutions can create trustworthy, sustainable, and innovation-oriented value beyond narrow economic performance.