Machine vision and formalist interpretation: Cultural implications of AI-based style recognition in oil painting

Abstract This study explores the intersection and conflict between formal and art-historical ways of thinking and machine vision powered by Artificial Intelligence in oil painting discovery and preservation. It demonstrates how visual signatures can be changed because of environmental, mechanical, and chemical degradations and be modified by interventions such as varnishing, gel and liquid cleanings, and structural reinforcement and how these developments AI perceives as stylistic choices rather than degradation artefacts. The article condemns algorithmic style recognition as a culturally and historically insensitive tool with canonically Western training data-based bias. It claims that the proper interpretation requires combining high-resolution imagery with rich conservation metadata, as well as human expertise to perform the necessary tasks in an otherwise hybrid system in which AI enhances rather than replaces formalist critique. Problems involving ethics, technology, and culture are described, and the roadmap ahead involves aspects that can involve predictive monitoring, the use of models that are trained with cultural awareness and mindfulness, and partnerships of conservators, art historians, and computer scientists to ensure authenticity that scales up in the way of analysis.