This graduation project explores the concept of Ubiquitous AI, an evolution of generative AI that moves beyond isolated, chatbot-based interactions toward AI embedded seamlessly within products and their ecosystems. Unlike the current AI assistant in Adobe Acrobat, which functions as a passive, panel-based chatbot requiring users to switch contexts, Ubiquitous AI envisions an ambient, anticipatory system that supports users without interrupting their workflows. Acting as a supervised yet autonomous cognitive co-pilot, the AI anticipates needs, surfaces assistance contextually, and performs actions when appropriate. The project investigates how interaction and visual design frameworks must evolve to support this shift toward more natural, human-centered AI experiences. Through research into generative AI capabilities and user needs, the study identifies key principles required to embed intelligence meaningfully into everyday tasks. The outcome positions Ubiquitous AI as a pathway to making complex technology more accessible, proactive, and aligned with real human workflows within Acrobat.