Augmented Decision-Making in Financial Markets: A Human–AI Co-Creative Approach
TL;DR
A co-creative framework in which the final decision combines AI-generated signals with human input is proposed, which shows that the allocation of decision authority between humans and AI can lead to different return–risk outcomes.
Abstract
This study examines the role of human judgment and artificial intelligence (AI) in financial decision-making. It proposes a co-creative framework in which the final decision combines AI-generated signals with human input. The relative influence of each component may change with market volatility, task complexity, and trust in the AI system. The framework is explored through a 252-trading-day simulation. Three stylized configurations are compared: an AI-only strategy, a human-only strategy, and a hybrid strategy that combines 60% AI input with 40% human input. The assessment includes cumulative and annualized returns, variance, maximum drawdown, and the Decision Quality Score (DQS). DQS is used as a comparative risk-adjusted indicator based on the mean–variance logic of modern portfolio theory. The simulation is intended as a proof of concept rather than a trading forecast. Under the assumptions used, the AI-only strategy generated the highest return and the highest DQS, although it was also associated with the greatest variance and the deepest drawdown. The human-only strategy produced the most conservative performance profile. The hybrid strategy fell between the two, preserving part of the AI model’s responsiveness while reducing some of its volatility. The study does not claim that one configuration is universally preferable. Instead, it shows that the allocation of decision authority between humans and AI can lead to different return–risk outcomes. Its main contribution is a structured framework for examining and comparing human–AI collaboration in financial decision-support settings. Received: 31 August 2025 | Revised: 3 April 2026 | Accepted: 2 July 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement Data sharing is not applicable to this article as no new data were created or analyzed in this study. Author Contribution Statement Liudmyla Bohrinovtseva: Conceptualization, Validation, Formal analysis, Writing – original draft, Writing – review & editing, Visualization. Olha Kliuchka: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing, Visualization, Supervision. Iryna Chunytska: Investigation, Writing – review & editing, Supervision. Olha Batrak: Methodology, Formal analysis, Investigation, Writing – review & editing, Visualization. Oleh Hustera: Methodology, Software, Validation, Formal analysis, Data curation.