Using DeepSeek’s reasoning-based AI model to support written feedback practices
Chenze Wu, Albert W. Li, Benjamin Luke Moorhouse · Assessing Writing · 2026
Generative Artificial Intelligence (GenAI) tools are increasingly being used to support written feedback practices for student writers. However, access to these tools may be restricted in certain contexts. The development and release of DeepSeek provide users with a different user experience through explicitly articulating the reasoning process behind its responses, which could potentially affect how diverse writers engage with it in their written feedback practices. In this review, we describe the key features of DeepSeek, contrast it with other GenAI tools, and explore its potential benefits and limitations for supporting student writers’ feedback practices. We also discuss ethical and practical implications, as well as future directions for using DeepSeek in writing assessment, instruction, and research. © 2026 The Author(s).