A Critical Analysis of the Reliability of Generative Models in Preserving Cultural Privacy: Towards Sustainable Digital Fashion Design Using Artificial Intelligence

This study aims to analyze the reliability of generative AI models in representing cultural privacy within the field of digital fashion design.A generative model was developed using a curated Saudi heritage database and tested on over fifty generated designs.The study adopted an applied evaluative methodology using an integrated toolset: a technical assessment by experts and a visual questionnaire for designers and specialists.Results indicated that identity-informed designs significantly outperformed non-informed ones in terms of cultural representation, attractiveness, wearability, and clarity of inspiration.Findings also confirmed that integrating heritage data substantially enhances the quality of visual outputs.The study presents a practical model for embedding cultural sensitivity into design algorithms and recommends the adoption of this methodology in other cultural contexts.