In the context of the artificial intelligence era, the application of artificial intelligence technology in modern brand design is increasingly widespread and is profoundly changing the design creation in this field. The purpose of this paper is to propose an intelligent image acquisition and automated brand design technology that utilizes AI technology to assist with brand design. The image acquisition and retrieval process of the online CBIR system is established, the average precision mean is used as the evaluation index of image retrieval, and the BoF model is constructed to realize RootSIFT feature extraction and image acquisition. Meanwhile, MultiStyleLogo GAN, a multi-style generative adversarial network, is designed to intelligently automate random output and generate brand design solutions through input sketches and style semantic descriptions to improve designers’ design efficiency. This paper utilizes technology to redesign the brand of A green plum wine in a province in China. Except for the pattern and font indicators, the average value of the newly designed brand image in each evaluation index of aesthetic design is greater than 4. The average values of the image aesthetics, cultural connotation, and emotional resonance indicators in the evaluation of brand narrative reach 4.25, 4.24, and 4.15, respectively, which achieve a high consumer satisfaction level.