This qualitative study examined the cognitive, behavioral, and emotional engagement of eight Chinese EFL undergraduates using generative AI for digital multimodal composing. Students employed various tools for seven goals, including text generation and translation, and applied four strategies to refine unsatisfactory outputs. Engagement was influenced by L2 proficiency, AI literacy, and contextual factors.
This autoethnographic study examines the author’s transition from traditional English-language classroom teaching to YouTube-based instruction. It analyzes how algorithmic milestones, learner engagement, and affective labour produce new forms of symbolic capital. The author argues that online teaching demands pedagogical adaptation and emotional resilience, while professional legitimacy increasingly derives from recognition within participatory communities rather than institutional hierarchies.
The author reviews Moorhouse & Wong’s 2025 book on generative AI in language teaching, criticizing its lack of an evidence base. Practical suggestions are deemed superficial, while risks such as bias, environmental costs, and language standardization are downplayed. The reviewer argues the book promotes a corporate agenda under the “AI for Good” banner, ignoring growing criticism of the technology.
A meta-analysis of 33 empirical studies (November 2022–March 2026) found a large overall effect of generative AI on EFL/ESL learning outcomes (g = 1.253). Strong effects emerged for language performance and affective outcomes, while self-regulation effects were non-significant. The evidence is overwhelmingly from tertiary education (87.9%) and rated as moderate certainty.