The purpose of this study is to examine the relationship between technology-supported writing instruction at the K-12 level and deep learning approaches and to understand the trends in this field. In the study, 12 articles selected from Web of Science, Scopus, ERIC and EBSCO databases were systematically analysed. The findings reveal that the methods differed in teaching writing, with the use of educational technologies varying according to educational level. While authentic learning environments were created with SVVR technology in primary school, artificial intelligence (AI)-supported assessment systems were used in secondary school and AI literacy and interdisciplinary approaches gained importance in high school. The diversity of research methods and the use of multiple data collection tools show that educational technology research has become more in-depth and comprehensive. The effectiveness of technology-supported writing processes varies according to students' cognitive development levels: in primary school, virtual and augmented-reality technologies improve writing performance; while in secondary school, AI-supported systems improve analytical thinking skills. In high school, NLP technologies and interdisciplinary applications strengthen higher-order thinking skills and writing abilities. At all levels, peer collaboration and feedback stand out as critical elements that support deep learning. The selection and application of educational technologies in accordance with students' cognitive development levels play an important role in determining the success of writing processes.