1　Introduction

Oral communication is the core goal and key pathway of foreign language learning, yet learners often avoid opportunities for expression due to apprehension and anxiety. Positive emotions can alleviate negative emotions, enhance learning engagement, and improve oral proficiency; however, no research to date has investigated intervention strategies for foreign language speaking enjoyment. In recent years, generative artificial intelligence has developed rapidly, and related studies have demonstrated its advantages in enhancing language skills, stimulating positive emotions, self-efficacy, learning motivation, and willingness to communicate. Based on this, the present study adopts a positive psychology perspective to experimentally examine whether generative AI-assisted speaking practice can enhance foreign language learners' speaking enjoyment levels and reveal its intervention mechanisms, with the aim of exploring the potential of this technology to optimize foreign language learning experiences and providing insights for the deep integration of foreign language education and emerging technologies.

2　Literature Review

This section reviews the "positive turn" in emotion research within the field of second language acquisition, noting that enjoyment research has become a leading direction in this domain. The research scope has gradually expanded from foreign language classroom enjoyment to skill-specific enjoyment in listening, speaking, reading, and writing, among which Jin & Qin (2024) developed a measurement scale specifically targeting Chinese university students' speaking enjoyment. Existing literature reveals that speaking enjoyment is influenced by factors such as activity interest, interactivity, and peer feedback, and can significantly improve the accuracy, fluency, and lexical richness of oral expression; however, intervention strategies for actively enhancing learners' speaking enjoyment have not yet been addressed. Meanwhile, generative artificial intelligence (GAI) technology has demonstrated multidimensional potential in foreign language teaching, including promoting language skill development, enhancing learning engagement and motivation, and stimulating positive emotions. These effects are highly correlated with the potential influencing factors of enjoyment emotions, providing empirical basis and research space for exploring GAI-assisted speaking practice in improving university students' foreign language speaking enjoyment levels and its underlying mechanisms.

3　Research Design

The basic information and grouping of 90 first-year student participants are described; the experimental and comparison groups were comparable in terms of age, gender ratio, and English proficiency. Research instruments included the "English Speaking Enjoyment Scale" (with only the "static enjoyment characteristics" subscale adopted) and GAI software based on large language models and speech recognition technology (using a情境自定义模式 with functions including pre-conversation preparation, real-time feedback and error correction during conversation, and post-conversation learning reports). The experiment lasted 12 days; the experimental group engaged in daily GAI dialogues on textbook topics for at least 30 minutes and regularly submitted written reflections, while the comparison group had no speaking tasks arranged. Both groups completed the scale at pre-test and post-test. Data analysis employed a mixed-methods approach: the quantitative component used descriptive statistics, paired-samples t-tests, and one-way analysis of covariance; the qualitative component adopted thematic analysis for coding, categorizing, and extracting themes from reflection materials, completed collaboratively by two authors to ensure research reliability.

4　Results

The intervention effects are presented through a combination of quantitative and qualitative methods. Quantitative results showed that the experimental group's post-test speaking enjoyment was significantly higher than pre-test (t=4.72, p<0.001, Cohen's d=0.82), with no significant change in the comparison group; covariance analysis further confirmed that, controlling for pre-test scores, the experimental group's enjoyment level was significantly higher than the comparison group (F=38.71, p<0.001, partial η²=0.31). The qualitative component, based on three reflections from the experimental group, identified seven core themes: the convenience, intelligence, and interactivity of GAI software created a positive practice environment; participants made progress in overall oral proficiency and dimensions such as fluency and accuracy; language knowledge including vocabulary, grammar, and pronunciation was enhanced; positive emotions including relaxation, accomplishment, happiness, and hope were experienced during dialogues; negative emotions such as tension and anxiety were alleviated; dialogue self-efficacy and oral proficiency improvement self-efficacy were enhanced; and motivation for speaking practice and overall English learning was stimulated.

5　Discussion

This section discusses the mechanisms and influencing factors of GAI-assisted speaking practice in enhancing university students' English speaking enjoyment. Quantitative results showed that the experimental group's speaking enjoyment was significantly higher than the comparison group at post-test, confirming the effectiveness of GAI-assisted practice. Qualitative analysis revealed seven key factors: positive speaking practice environment (scaffolding functions of GAI, immediate feedback and error correction, low-pressure error-tolerant characteristics, rich language expression, etc., forming learning affordances); oral proficiency development; improvement in language knowledge and application ability (satisfying psychological needs, enhancing foreign language ability perception and self-concept); interactive effects of positive emotion stimulation and negative emotion alleviation (emotional reciprocity creating conditions for enjoyment); self-efficacy enhancement (strengthening growth mindset, intrinsic motivation, and academic perseverance, increasing sense of control); and motivation strengthening (primarily intrinsic motivation based on interaction and growth). Complex interactive relationships exist among these factors, jointly promoting speaking enjoyment. The study combines new developments in educational technology with speaking enjoyment research, indicating that GAI-assisted speaking practice can produce multifaceted positive impacts, providing prospects for digital language teaching and positive psychology research, and recommends that foreign language teachers enhance digital literacy and effectively integrate it into teaching to promote teacher-student knowledge and emotional development.

6　Conclusion

This section elaborates on the effectiveness of GAI-assisted speaking practice in enhancing foreign language speaking enjoyment, identifying positive practice environment, oral proficiency development, language knowledge improvement, positive emotion stimulation, negative emotion alleviation, self-efficacy enhancement, and motivation strengthening as potential influencing factors, while acknowledging that the study represents only a preliminary examination and that intervention schemes require further validation in different contexts and populations. It is recommended that future research expand to diverse speaking emotions and proficiency studies to deeply explore the application prospects of GAI in foreign language education.