This study explores the application of generative artificial intelligence (AI) to the design of facades for high-rise office buildings. It is motivated by the increasing demands for speed and quality in architectural design, as well as the need to comply with contemporary aesthetic and functional standards. Traditional design methods, which rely on manual labor, often fail to keep pace with market dynamics, making the adoption of AI technologies a promising alternative. The study evaluates the capabilities of the Exactly.ai platform for automating facade design, assessing its effectiveness through a comparative analysis with other neural networks (Midjourney, Stable Diffusion) across key parameters: style replication accuracy, adaptability to project requirements, and generation speed. A custom model was trained on a dataset of 48 images of office buildings, and its performance was tested by generating facades from textual prompts and sketches. Results indicate that Exactly.ai achieves 68% compliance with specified stylistic and functional criteria, demonstrating high processing speed (2.5 s/image) and strong realism (88%). However, limitations were identified in detailing high-rise structures and material representation accuracy. The study confirms the potential of AI in architectural design, particularly during early-stage conceptualization. This study positions AI as a tool that augments, rather than replaces, the architect.