Artificial intelligence is becoming part of lesson planning, tutoring, feedback, assessment, learning analytics, and institutional decision making. However, growth in AI-assisted teaching does not automatically lead to fair educational outcomes. This study maps research on AI-assisted teaching and examines how educational equity appears within the field. Records published from 2010 to 2025were retrieved from Google Scholar and Scopus during July and August 2026. After duplicate removal and screening, 1,346 journal articles and reviews were retained. Bibliometric procedures included annual publication analysis, thematic coding, country and authorship analysis, keyword co-occurrence review, and an equity-focused content classification. The results show rapid publication growth after 2022, with 553 records appearing in 2025 and the first eight months of 2026. Research is concentrated in instructional design, personalized learning, assessment, and teacher support, while equity, inclusion, ethics, and governance form a smaller but growing area. High-income countries account for 67.8% of corresponding authorship, whereas low-income countries account for only 1.0%. Among 318 records with an explicit equity focus, disability and accessibility, socioeconomic access, language inclusion, and algorithmic fairness receive the most attention. The paper argues that future work should connect technical performance with access, representation, teacher agency, data protection, and locally meaningful outcomes. It also proposes a research agenda for more inclusive datasets, stronger comparative designs, and human-centered governance in AI-assisted teaching