This paper introduces AI Visibility as a formal discipline concerned with how information is authored, structured, and emitted so it can be reliably ingested, retained, and recalled by large language models. It presents a canonical definition and a theoretical framework describing upstream conditions that influence learnability, attribution stability, and semantic coherence during model training. Earlier scientific publications by the author appear under the name J. Mas in nuclear physics collaborations.
Joseph Mas (2026) studied this question.