AI Dependency Syndrome: Exploration and Identification via Blockchain-Based Machine Learning Approach

Abstract

In the present era, the pervasive adoption of Artificial Intelligence (AI) has resulted in increasing human reliance on automated systems, giving rise to what we define as AI Dependency Syndrome (ADS). ADS is characterized by a gradual decline in human creativity, critical thinking, and problem-solving abilities, necessitating systematic investigation. This study proposes a quantitative framework for identifying and predicting ADS using an ensemble of ten machine learning classifiers. To enhance data integrity, transparency, and privacy, blockchain technology is integrated into the analytical pipeline. Experimental results show classifier accuracies ranging from 78.45% to 92.67%, demonstrating notable performance variation across models. A comparative analysis identifies the most effective classifiers for ADS prediction. The proposed blockchain-enhanced machine learning framework provides reliable insights into AI dependency patterns and supports the development of informed mitigation strategies. These findings contribute toward promoting a balanced, human-centric integration of AI while minimizing its potential adverse cognitive and societal impacts.