This paper proposes a new framework for artificial intelligence (AI) enhancement inspired by behavioral economics concepts. The proposed framework targets human decision-making patterns, integrating cognitive bias into AI algorithms to improve their ability to model and predict human behavior. Loss aversion, reference dependence, and time inconsistency are suggested as the first relevant concepts that can be structured within an enhancement approach for AI algorithms. For each identified mechanism, the paper provides mathematical formulations and implementation strategies that seek to preserve both computational efficiency and psychological validity. This research highlights healthcare, banking, transportation, and energy management as the main sectors for practical applications of behavioral-AI, which align with human decision-making proclivities, as these sectors exhibit the highest disparity between AI’s technical proficiency and its capacity to anticipate and engage with human behavior. We hope that this work stimulates the development, testing, and evaluation of new methodologies, frameworks, and evaluation metrics that will help us improve AI alignment with human decision-making behavior.