It introduces the development history of action research, emphasizing the importance of teachers' self-reflective inquiry in real teaching situations for improving teaching practices and enhancing teaching effectiveness. With the development of large language models and generative artificial intelligence technologies, these technologies are transforming traditional foreign language teaching models and providing new opportunities for school education. The "College English Teaching Guide (2020 Edition)" highlights the significant role of modern technology in English teaching and promotes the deep integration of technology and teaching. In the context of GAI-empowered teaching, teachers need to transition from knowledge transmitters to designers of collaborative teaching. Based on ecological affordance theory, this study constructs a GAI-empowered "triadic six-step" human-machine collaborative writing teaching model. By collecting both process and outcome data, it adjusts teaching plans to promote the improvement of students' writing abilities.

1. Theoretical Foundations and Implementation Plan for GAI-Empowered College English Writing Instruction

This section discusses the challenges of providing personalized and timely feedback in college English writing instruction and introduces the development of Automated Writing Evaluation (AWE) systems. It points out the limitations of traditional AWE systems, such as the inability to recognize handwritten essays and the lack of personalized feedback. Subsequently, the section highlights the advantages of GAI intelligent platforms in the field of second language writing assistance, such as supporting multimodal file uploads and providing comprehensive writing optimization frameworks. Drawing on ecological affordance theory, the section elaborates on the importance of the interaction between learners and their environment and proposes three core phases: "perception," "interpretation," and "action." Based on this, it constructs a GAI-empowered "triadic six-step" human-machine collaborative teaching model for college English writing. This model involves triadic interaction among teachers, learners, and GAI, as well as six dynamically iterative teaching steps.

2. Action Research on GAI-Empowered College English Writing Instruction

This chapter introduces an action research study on the application of Generative Artificial Intelligence (GAI) in college English writing instruction. The research adopts a two-round spiral action research paradigm, consisting of four core steps: "planning, action, observation, reflection." The first round of action research was implemented in three non-English major freshman classes at Yanshan University. Writing content was determined based on the "New Horizon College English Reading and Writing Course," and writing tasks were differentiated according to students' English proficiency levels. Students received writing feedback through a GAI platform and engaged in self-reflection and revision. The study found that students could perceive the writing assistance provided by GAI and improve their writing quality accordingly. However, some issues were also identified, such as inappropriate suggestions in GAI feedback, students' over-reliance on GAI feedback, and a lack of persistent inquiry behavior.

In response to the issues from the first round, the second round of action research adjusted the teaching plan. These adjustments included guiding students to critically analyze GAI feedback, modifying writing tasks, optimizing requirements for GAI feedback, and constructing a personalized reflection journal framework. In the second round of practice, students demonstrated a stronger tendency for personalization when using the GAI platform, being able to select and combine different platform functions according to their own needs. The quality of students' second drafts significantly improved, and they were able to utilize GAI feedback more effectively for revision. Students generally were able to reflect based on the GPS framework, outlining their gains, problems, and improvement plans.

The study emphasizes the potential of GAI in college English writing instruction but also points out students' over-reliance on GAI feedback and deficiencies in critical thinking abilities. The research proposes strategies for optimizing GAI-empowered teaching, such as cultivating students' critical thinking, metacognitive monitoring skills, and AI ethics literacy, as well as establishing diversified feedback mechanisms. Through two rounds of action research, the researchers explored the application of GAI in college English writing instruction and put forward corresponding suggestions for teaching improvement.

3. Conclusion

This chapter summarizes the two rounds of directed action research. By integrating the three elements of ecological affordance theory—"perception, interpretation, action"—it constructs a GAI-empowered "triadic six-step" human-machine collaborative teaching model for college English writing. The study finds that this model effectively improves students' writing quality, enhances their behavioral and cognitive engagement, reduces learning anxiety, increases emotional investment, and establishes a virtuous cycle of "technological affordance—pedagogical transformation—emotional identification." In GAI-empowered teaching, teachers play the role of an "affordance transformation catalyst." They need to understand GAI logic, guide students to experience adjustable difficulty levels, cultivate critical thinking and metacognitive abilities, and transform GAI feedback into sustainable writing strategies. Acting as scaffolds for perception, interpretation, and action, teachers infuse the collaborative co-cultivation model with professional judgment and educational warmth. Continuously optimized teaching models and hybrid feedback mechanisms will open new horizons for foreign language writing instruction, activating new potential for teaching and learning.