The acceleration of the process of urban identity development is becoming an increasingly important task in the context of the growing need for cities to be vivid and memorable. The article considers an approach to solving this problem using artificial intelligence and generative design technologies. The main focus is on creating a system that, based on data from open sources, such as Wikipedia, can automatically generate visual images reflecting the unique cultural, historical and geographical features of a particular locality. The project architecture includes client part on React, server part on FastAPI, PostgreSQL database and generative module using Fusion Brain API. Experimental studies of the system have shown its applicability for creating visual images of various cities in different styles and colors. Comparative analysis with existing solutions showed that the system has the advantages of adaptability, openness, localization and scalability. The solution is aimed primarily at city marketing specialists and demonstrates the potential of a generative approach in automating visual branding processes for any communities and territories.