Dark Pattern Detection Using Machine Learning

Dark Pattern Detection Using Machine Learning

V. Bharathi
Department of Artificial Intelligence and Data Science,
St. Joseph's Institute of Technology College, Chennai - 600119, Tamil Nadu, India.

Abstract

Dark patterns are deceptive design techniques used in digital interfaces to manipulate user behavior, often leading to unintended actions. These unethical design choices negatively impact user autonomy, causing financial loss, privacy violations, and psychological distress. This paper presents a machine learning approach to detecting dark patterns in online platforms. By leveraging deep learning models, natural language processing (NLP), and computer vision techniques, our system identifies misleading UI/UX elements that unfairly influence user decisions. The proposed method achieves high accuracy in detecting dark patterns such as forced continuity, hidden costs, and disguised advertisements. Furthermore, this research emphasizes the ethical implications of dark patterns and the necessity for automated solutions in mitigating their prevalence. Our approach contributes to the broader discourse on responsible design and regulatory enforcement.

Keywords: Dark patterns, machine learning, deep learning, deceptive design, UI/UX, user manipulation, ethical AI, regulatory frameworks