Abstract
<jats:title>Abstract</jats:title> <jats:p>Academic writing, a key area of applied linguistic research and pedagogy, has been a focus of artificial intelligence (AI) since AI began. The overlapping analytic and evaluative branches of AI in academic writing, natural language processing (NLP) and automated writing evaluation (AWE), have developed significantly since that time, from top‐down tools accounting for discrete features to iterative large model tools shaped by AI and human collaboration in real time. Three challenges have persisted throughout those developments: representation (what language and other data are represented in AI parameters and models), characterization (what human and AI judgments and classifications are made), and engagement (how researchers and educators critically address and use AI in academic writing). This entry will offer an overview of the evolution of AI in academic writing and will discuss these challenges and their implications before closing with ethical considerations.</jats:p>