IntroductionThe first decade of the 21st century has witnessed significant development and widespread application of computer technology, the internet, and online communication technologies globally (Wang et al., 2025a). These advancements have fundamentally transformed the way people work, communicate, socialize, and acquire knowledge (O’Dowd, 2009; Wang et al., 2023; Jing et al., 2023). Since the COVID-19 pandemic, there have been increasing efforts to integrate digital technologies and social media more effectively into foreign language education (Lin et al., 2023). Technology-enhanced language learning (TELL) has exhibited substantial educational potential across diverse educational contexts, including massive open online courses, flipped learning, and self-directed learning that leverages various e-learning resources and tools (Jin et al., 2025a; Jing et al., 2024a). In contrast, traditional language instruction has traditionally focused on grammar exercises and mechanical worksheet completion. However, the integration of e-learning resources and tools such as blog writing, online movies, and musical pieces can effectively address the limitations of conventional education and enrich the overall learning experience for learners (Lai, 2015; Levy, 2009).Foreign language learning often faces various challenges, with foreign language anxiety (FLA) serving as a notable barrier among them (Alshumaimeri & Alhumud, 2021). Within the realm of Foreign language learning, FLA poses a unique challenge. Language learners are expected not only to grasp the linguistic complexities of the foreign language but also to explore the cultural nuances and literary devices inherent to another culture (Lin & Yu, 2025b). The combination of linguistic challenges and cultural insights can exacerbate FLA, leading to frustration and disengagement among students (Cruz & Anderson, 2024). Recent technological advancements provide promising avenues for addressing FLA and enhancing the foreign literature learning experience.It is widely acknowledged that technology has brought about transformative changes in the methods of learning foreign languages (Jin et al., 2025b). However, the issue of whether relying on technology can enhance the language learning experience remains unresolved. This relationship demonstrates significant complexity. There are still several aspects of TELL that require further investigation. Technology, regardless of its form, does not automatically result in improved learning outcomes (Warschauer & Kern, 2000). To achieve better language learning outcomes and enhance language learners’ experiences under TELL, it is essential to conduct a reasonable assessment and rational judgment of the impact of technology on learner anxiety in the field of language learning.Although many empirical studies have investigated the impact of technology on FLA, comprehensive reviews remain remarkably scarce, with only a thematic review by Ma et al. (2022) and a brief research synthesis by Aydın (2018). Our study significantly advances this field by employing a systematic literature review (SLR) combined with grounded theory, an innovative methodological approach previously unused in this research domain. Through rigorous analysis of 99 empirical articles, more than any of the previous reviews, our study not only provides a more comprehensive synthesis but also develops novel theoretical frameworks to better understand the complex relationship between TELL and FLA. Specifically, our grounded theory approach enables us to systematically identify and categorize both anxiety-alleviating and anxiety-exacerbating mechanisms of various technologies, offering unprecedented granularity in understanding this crucial educational issue. Specifically, we address the following research questions:RQ1: What core dimensions are included in the research frameworks of empirical studies on the impact of technology on foreign language learning anxiety?RQ2: What are the mechanisms by which technology exacerbates or mitigates foreign language learning anxiety?RQ3: Based on the review of existing literature, what potentially valuable research agendas can be identified for future studies on the application of technology in reducing foreign language learning anxiety?Literature reviewForeign language anxietyFLA is a critical psychological factor that significantly impacts foreign language teaching and learning. Research has consistently shown that FLA hinders language acquisition and serves as a key predictor of learner performance (Altamimi, 2023; Yuan et al., 2025; Li & Li, 2024). Recent studies have explored its mechanisms, effects, and mitigation strategies, highlighting the importance of addressing FLA to enhance language education (Wang & Hui, 2024; Zhang & Lai, 2024). For example, Oruç and Demirci (2020) reported that FLA negatively impacts language achievement but that student engagement can mediate this effect. They emphasized fostering engagement to reduce FLA, although their cross-sectional design limited causal inferences. Chahrazad and Kamel (2022) highlighted the role of supportive classroom environments in managing speaking anxiety, recommending communicative activities and relaxation techniques. However, their strategies lacked specificity for different proficiency levels. Technology integration has also been explored as a solution. Gok et al. (2023) demonstrated that online flipped classrooms significantly reduce classroom and reading anxiety, highlighting the potential of technology-enhanced learning. Nevertheless, individual preferences for online learning and their impact on other language skills remain underexplored. Zhou et al. (2023) further revealed that FLA negatively moderates the relationship between language competence and willingness to communicate, emphasizing the need to address anxiety to improve communication.Furthermore, FLA is a complex, multidimensional construct that interacts dynamically with foreign language enjoyment in technology-mediated learning environments (Dewaele, 2024; Yeşilçınar et al., 2025; Rezai et al., 2025; Wang et al., 2024a; Zare et al., 2022; Wang et al., 2025b). Research has established robust measurement tools such as the Short-Form Foreign Language Enjoyment Scale and Foreign Language Classroom Anxiety Scale (Staedtler et al., 2025; Pan et al., 2025; Huang & Zhang, 2024), which have demonstrated cross-cultural validity in diverse educational contexts (Yeşilçınar et al., 2025; Dewaele, 2022). Studies consistently identify key predictors of FLA, including learners’ self-perceived proficiency, multilingual background, and trait emotional intelligence, while highlighting gender and cultural variations in emotional responses (Dewaele & MacIntyre, 2024; Dewaele et al., 2025; Lu, 2025). Despite these advances, however, the literature presents conflicting findings regarding the role of technology in either exacerbating or alleviating FLA.The pedagogical dimension of FLA research has yielded important insights into teacher-related factors, with multiple studies confirming that instructor enthusiasm, emotional support, and scaffolding techniques significantly reduce learner anxiety while enhancing enjoyment (Dewaele & Li, 2021; Derakhshan et al., 2025; Mettewie et al., 2024). The unexpected global shift to emergency remote teaching during the COVID-19 pandemic provided a natural experiment that revealed the dual effects of technology on FLA—reducing performance anxiety for some learners while introducing new technostress factors for others (Derakhshan et al., 2025; Resnik & Dewaele, 2023). Notably, virtual environments appear to disrupt the typical inverse relationship between foreign language enjoyment and FLA observed in physical classrooms, suggesting fundamental differences in emotional dynamics across learning modalities (Resnik & Dewaele, 2020). These findings point to significant gaps in our understanding of how different technological interfaces mediate affective experiences in language learning.Emerging research on technological interventions has produced promising but inconsistent results. While immersive technologies such as virtual reality (VR) help reduce public speaking anxiety (Kaplan-Rakowski & Gruber, 2023; Zhao et al., 2024; Diert-Boté & Moncada-Comas, 2024), and flipped classrooms show efficacy in addressing reading anxiety (Gok et al., 2023), other areas remain understudied. Writing anxiety, for example, appears more responsive to self-efficacy interventions than to technological solutions alone (Woodrow, 2011). Furthermore, critical questions persist regarding individual differences in responses to technology-mediated instruction, with preliminary evidence suggesting that factors such as age of language acquisition, frequency of target language use, and socialization patterns may moderate these effects (Dewaele et al., 2025; Zhou et al., 2023).The role of technology in foreign language learningIn today’s fast-changing educational environment, technology has become a key driver of innovation, transforming traditional teaching methods and creating more effective learning opportunities (Lin et al., 2023; Liu et al., 2024a; Dai & Wang, 2024; Wang et al., 2025c). Integrating technology can boost engagement and improve language acquisition. For example, multimedia tools enhance comprehension and retention, although challenges such as distraction due to excessive use and unequal access to infrastructure must be addressed (İnceçay & Koçoğlu, 2017; Xu, 2024). Automatic speech recognition provides instant feedback to improve pronunciation, while chat-based interactions increase the volume and complexity of learners’ language production (Golonka et al., 2014). Artificial intelligence (AI) tools such as ChatGPT supplement and elevate conventional English as a foreign language teaching approaches (Mohamed, 2024), offering practical solutions to prepare students, such as English as a foreign language learners in Japan, to communicate confidently in English and participate in the global landscape (Busso & Sanchez, 2024).Video games also play a notable role in language learning, particularly in enhancing vocabulary and pronunciation. As engaging leisure activities, video games create a supportive and motivating environment that is ideal for foreign language acquisition (Winaldo & Oktaviani, 2022). Similarly, technology-mediated communication enables students to practice meaning negotiation and develop pragmatic skills in real-world contexts, while exposing them to diverse genres, registers, and youth language that were previously difficult to access without living abroad (Kern, 2014). Additionally, the immersive, interactive, and imaginative features of VR foster heightened interest and active participation, especially in speaking, listening, and cultural practice (Li et al., 2021).However, the integration of technology is not without challenges. Issues such as unequal access, overreliance on digital tools (Zhao & Lai, 2023), and resistance to adopting new teaching methods must be addressed. Translanguaging, for example, can increase student participation and create a more inclusive classroom environment, but it requires continuous support for educators to implement it effectively (Xu & Fang, 2024). Moreover, technology should not be viewed as a universal solution or an end goal but rather as a tool to achieve specific learning objectives (Chun et al., 2016).In conclusion, although technology-driven language learning often matches or exceeds the results of human-led methods (Zhao, 2004), its successful implementation depends on addressing challenges and balancing its use with pedagogical strategies. By doing so, educators can harness their potential to create dynamic, inclusive, and effective learning environments.Comparative analysis of literature reviews on technology and foreign language anxietyThis review employs the SLR method, which offers a more rigorous and standardized approach than the thematic review by Ma et al. (2022) and the brief research synthesis by Aydın (2018). While Aydın’s study synthesized 28 empirical articles (2009–2016) using qualitative analysis, it did not explore technologies related to AI; instead, it focused on computers and mobile devices and lacked coverage of emerging technologies (e.g., social media, AI). Similarly, Ma et al. (2022) conducted a meta-analysis of 24 studies (2016–2021) and reported that 46% of interventions reduced FLA, whereas 54% of interventions had no significant effect; additionally, moderators such as technology type and exposure duration were examined. However, both prior reviews failed to link specific technologies systematically to FLA components. In contrast, the present SLR encompasses a broader timeframe, incorporates 99 empirical and peer-reviewed studies in prominent academic databases, and applies grounded theory to code anxiety-alleviating mechanisms, thereby enhancing methodological robustness and theoretical depth.The current review significantly advances the field by providing a comprehensive FLA analysis that explores all the components - communication apprehension, fear of negative evaluation, and test anxiety - and directly links them to technology affordances (e.g., VR for immersive speaking practice, AI for personalized feedback). Unlike earlier works, this study highlights the teacher’s role in scaffolding technology integration (e.g., hybrid classrooms) and traces the evolution of technological impacts from Web 1.0 tools to AI-driven solutions, revealing a trend of iterative anxiety reduction. Furthermore, it addresses critical gaps in prior reviews by examining the dual roles of social media in exacerbating or mitigating FLA, a dimension overlooked by both Aydın (2018) and Ma et al. (2022). By proposing a technology-anxiety mitigation model, this study establishes theoretical linkages between tool features (e.g., anonymity, instant feedback) and FLA dimensions, thus offering a framework for future research.The policy and practical implications of this review are substantial. While Aydın (2018) and Ma et al. (2022) offered general insights, this SLR delivers actionable strategies tailored to specific anxiety types and learner demographics. For example, VR is recommended for reducing speaking anxiety, and AI chatbots for alleviating fear of negative evaluation, with these strategies bridging the gap between theoretical findings and classroom practice. By synthesizing a wider array of technologies and incorporating longitudinal trends, this review not only consolidates past findings but also sets a new benchmark for rigor and applicability in the field of technology-mediated language learning.Methodology and materialsMethodologyThis study employs SLR to examine the impact of technology on language learning anxiety and analyzes both exacerbating and alleviating mechanisms to identify future research directions. The SLR provides a rigorous framework for defining research questions, retrieving literature, and evaluating studies while minimizing bias (Diekemper et al., 2015; Paul et al., 2021). Its standardized approach is particularly suited to social science research, as it integrates qualitative evaluation metrics (Shlonsky et al., 2011; Deacon et al., 2023) and aligns with our goal of synthesizing diverse findings in this field (Gough et al., 2017; Jing et al., 2024b; Chen et al., 2025).Our methodology followed a structured three-phase process (Fig. 1): Preferred reporting items for systematic reviews and meta-analyses (PRISMA)-guided literature screening and coding (Dai et al., 2024; Page et al., 2021; Lin & Lan, 2015), establishing a dual coding system for both structured metadata (publication characteristics) and unstructured textual data (anxiety mechanisms); framework analysis of core dimensions (Empirical landscape of technology and foreign language anxiety research (RQ1)); and grounded theory analysis of mechanisms (“Grounded analysis for mechanisms of technology-related foreign language anxiety (RQ2)“, culminating in future research directions (Future agenda (RQ3)).Fig. 1: Overview practical logic of research.Columns represent sequential research phases, with bold section numbers indicating corresponding paper sections.Full size imageLiterature selection processBy the PRISMA guidelines, SLR typically requires three or more databases as sources of literature (Page et al., 2021). To comprehensively obtain the literature needed for our study, we referenced existing SLR studies in the field of educational technology (Radianti et al., 2020; Luo et al., 2021) and selected four English-language databases: EBSCO, Ei Compendex, Scopus, and Web of Science. These databases extensively cover leading educational journals, and their inclusion criteria ensure that the literature within them generally meets the quality requirements for SLR (Chen et al., 2024; Jing et al., 2024b; Liu et al., 2025; Ma & Ismail, 2025).In setting the search criteria, our study referred to keyword lists from literature in the fields of technology and language learning anxiety to identify appropriate keywords. For example, the final subject search query in Web of Science was TS = ((“language learning” or “language acquisition” or “linguistic education” or “language study” or “language mastery” or “language education” or “language training” or “language improvement”) AND “anxie*” AND “technolog*”). The search timeframe was set from January 2004 to December 2024, with the cutoff date being December 31, 2024.Figure 2 presents the systematic literature selection process following PRISMA guidelines (Page et al., 2021). Initial searches across four rigorously selected databases—Web of Science (112 records), Ei Compendex (107 records), Scopus (222 records), and EBSCO (272 records)—yielded 713 total records. After 205 duplicates were removed, 508 unique records underwent title and abstract screening using predefined criteria derived from prior SLR methodologies (Radianti et al., 2020; Luo et al., 2021). This phase excluded 137 records, leaving 371 for full-text review. The final inclusion process further excluded 272 records due to irrelevance or low thematic correlation, resulting in 99 high-quality articles for analysis. This stringent process ensures the review’s comprehensiveness while meeting SLR quality standards (Chen et al., 2024), as detailed in “Methodology and materials”.Fig. 2: PRISMA flowchart for including studies to review.Data sources include Web of Science, Scopus, EI Compendex, and EBSCO (APA PsycInfo & ERIC).Full size imageThe sample of 99 empirical articles in our study is both reasonable and methodologically sound for several reasons. First, the field of technology-mediated foreign language anxiety remains a niche yet rapidly evolving domain, meaning that the available literature is inherently limited compared with more established research areas. Our stringent inclusion criteria—prioritizing empirical studies with direct relevance to FLA and technology—further refined the corpus to ensure analytical depth over breadth.This sample size aligns with recent systematic literature reviews in related fields. For example, Radianti et al. (2020) examined virtual reality in education using only 59 articles, while Luo et al. (2021) conducted an SLR on a comparable topic with 149 articles. Unlike large-scale bibliometric analyses (Ma & Wang, 2025; Lin & Yu, 2025a; Lin & Yu, 2024; Zhou et al., 2025), which prioritize broad trend detection, our study adopted a grounded theory approach (Charmaz & Thornberg, 2021) to uncover underlying mechanisms. This approach required in-depth qualitative engagement with texts, exemplified by our identification of 35 open codes related to anxiety exacerbation—a level of granularity impossible to achieve with larger but less curated datasets. Thus, our sample size reflects a deliberate trade-off: sacrificing quantitative generalizability for rich, contextually grounded insights into how technology influences FLA.Following our rigorous selection protocol, publication trends revealed significant growth in research on the role of technology in language learning anxiety, particularly following the COVID-19 pandemic (2020-2022) and the emergence of ChatGPT (2022-2024). Geographically, China contributed the majority of studies, followed by Turkey, the U.S., and Malaysia. Our analysis identified Computer Assisted Language Learning, Education and Information Technologies, and System as the dominant journals publishing on this topic. These findings provide contextual data for better systematic analysis.Manual screeningThe initial search often yields articles that appear relevant but are not pertinent. Therefore, manual screening is essential to ensure that only literature relevant to the research topic is included (Pollock & Berge, 2018; Wang et al., 2024b).To conduct the manual screening process efficiently and scientifically while maintaining the quality of included literature, the research team established the following criteria (Table 1).Table 1 Inclusion criteria.Full size tableOur manual screening employed a rigorous two-phase approach with clearly defined inclusion criteria to ensure methodological consistency. For research content, we exclusively selected studies that centrally investigated the dual impact of technology on FLA, including exacerbating factors (e.g., cognitive overload), alleviating mechanisms (e.g., psychological safety), and underlying theoretical explanations. Studies that merely mentioned the topic peripherally were excluded. With respect to literature quality, we included only original research articles with full-text availability from reputable academic databases, complete bibliographic metadata, and standardized peer-review certification. The screening process involved independent dual-reviewer assessments followed by consensus-based full-text evaluations, ensuring alignment with PRISMA standards while maintaining a focus on the study’s core research objectives. This systematic approach guaranteed both the relevance and academic rigor of the selected literature.Analytical codingFollowing the completion of the literature screening, we conducted a comprehensive coding analysis of the 99 ultimately included research articles (all original research papers) included in this study. Based on systematic literature review methodology (Luo et al., 2021), we developed a four-dimensional coding framework: metadata, research context, technology types, and research objectives. This structured classification system (detailed in Table 2) supports both quantitative statistical analysis while preserving in-depth insights from qualitative research.Table 2 Lists of codes for the analysis of selected articles.Full size tableThis coding scheme offers three key advantages. First, systematic analysis is ensured through multidimensional data extraction (methodological characteristics, technology types, and theoretical frameworks). Second, it effectively identifies patterns and contradictions in the literature by combining quantitative statistics (such as frequency analysis of technology types) with qualitative thematic analysis. Finally, it strictly adheres to coding standards for systematic reviews, laying the foundation for subsequent grounded theory analysis. Particularly noteworthy is how the coding of technology’s “double-edged sword” effects in the research objectives dimension directly supports this study’s core theoretical contribution—revealing the dialectical mechanisms through which technology influences language anxiety.Following the development of the coding framework, an independent coder with expertise in educational technology—a Ph.D. student—was enlisted to code the data in parallel, thereby enhancing the study’s reliability (Gaur & Kumar, 2018). Discrepancies in coding were resolved through collaborative discussions, while significant disagreements were adjudicated by consulting a recognized expert in the field. The coding procedure was completed over two weeks.Empirical landscape of technology and foreign language anxiety research (RQ1)This section addresses RQ1 by synthesizing the core conceptual, contextual, and methodological characteristics of existing empirical studies. It provides a descriptive mapping of the research landscape and highlights foundational gaps that limit theoretical and pedagogical advancements in the field.Theoretical perspectives on and learner populations in technology-related foreign language anxiety researchThe current research landscape reveals several critical gaps in understanding the impact of technology on FLA. As shown in Fig. 3(A), only 39% of the studies specifically examined anxiety, suggesting that FLA is often treated as a secondary consideration despite its recognized complexity and frequent comorbidity with other psychological factors. The theoretical foundations appear particularly weak, with just 10 studies (Fig. 3C) employing established frameworks-predominantly the technology acceptance model (6 studies, e.g., Li et al., 2019; Cakir & Solak, 2014; Dizon & Thanyawatpokin, 2021) and willingness to communicate (4 studies, e.g., Lee & Drajati, 2019)—whereas 90.91% lack explicit theoretical grounding. This theoretical deficit indicates that the field remains in its nascent stages and lacks comprehensive models to explain how technologies influence the cognitive, affective, and physiological dimensions of FLA.Fig. 3: Characteristics of the included studies.(A) Proportion of research specializing in anxiety. (B) Participants’ grade level. (C) Applied theoretical frameworks in articles. Panel A shows anxiety research specialization, Panel B displays participant demographics, and Panel C indicates theoretical framework application gaps in current literature.Full size imageFurthermore, Fig. 3(B) illustrates a concerning research bias toward higher education populations, which potentially limits our understanding of developmental differences in technology-FLA interactions. While adult learners’ psychological maturity may facilitate technology use, this focus neglects crucial periods of language learning anxiety formation in younger students. The combined findings suggest an urgent need for theory-driven research integrating established FLA models with technology frameworks, developmental studies across educational levels, and more systematic investigations of how specific technological features mediate various FLA manifestations. Without addressing these gaps, the field risks developing technologies that fail to account for the full spectrum of learners’ anxiety experiences.The intersection of technology types and skill-specific foreign language anxietyAs illustrated in Fig. 4, our review shows that studies on AI technologies, accounting for 29 (29.29%) items, focused on applications including conversational AI chatbots, such as ChatGPT (Tram et al., 2024), automated writing evaluation (Sari & Han, 2024), speech-to-text recognition (Shadiev et al., 2024) and affective computing (Chen & Lee, 2011), which address emotional factors such as motivation, boredom (Kruk, 2022), and self-efficacy (Sari & Han, 2024) in language learning. Studies on VR and augmented reality (AR) technologies, comprising 15 (15.16%) items, emphasized immersive experiences through virtual reality applications such as VirtualSpeech (Alsaffar, 2021), high-immersion virtual reality (Ding, 2024), and augmented reality tools such as AR filters (Zhu et al., 2024), which enhance speaking and presentation skills. Studies on online and mobile learning, which accounted for 55 (55.56%) items, covered a wide range of platforms and tools, including learning management systems, such as Canvas (Yaprak, 2022), mobile apps, such as Duolingo (Neuschafer, 2023) and Rosetta Stone (Bai, 2024); and social media tools, such as Telegram (Çakmak et al., 2023) and YouTube (Jin, 2023), which support various language skills, such as speaking, writing, reading, and listening. This small proportion of VR and AR research might be related to the low accessibility and high costs of this technology.Fig. 4: Relationship between technology studied and language skills.Data reveals disproportionate focus on productive skills (speaking/writing) versus receptive skills (reading/listening), and emerging technologies (VR/AR) showing lower research coverage.Full size imageThe analysis of technological applications in language education reveals several critical insights regarding their relationship with FLA. The current research landscape demonstrates a predominant focus on AI technologies (29.29% of studies), which show particular efficacy in addressing FLA through multiple mechanisms. Conversational AI tools such as ChatGPT provide low-stakes speaking practice environments, while automated writing evaluation systems offer nonjudgmental feedback to alleviate writing anxiety. Significantly, affective computing applications directly target emotional factors such as motivation and self-efficacy, which mediate FLA experiences.The research reveals a strong emphasis on speaking skills (41 applications across all technologies), reflecting their well-documented anxiety-inducing potential, with VR and AR systems (15.16% of studies) showing particular promise for creating immersive, controlled exposure environments despite current accessibility challenges. The predominance of online and mobile platforms (55.56%) highlights their practical advantages for anxiety reduction through flexible, accessible learning formats. However, the relative neglect of writing (9%) and grammar instruction in technological applications indicates a significant gap in addressing these equally anxiety-provoking skill areas. These findings collectively suggest that while existing technologies demonstrate substantial potential for FLA intervention, their application remains uneven across language domains. Future development should prioritize: more balanced technological solutions across all language skills, cost-effective VR and AR implementations for anxiety-sensitive pedagogies, and systematic integration of affective computing principles to address the multifaceted nature of FLA in diverse learning contexts.Methodological characteristics and trends in technology-related foreign language anxiety studiesA statistical analysis was conducted on the types of literature, with a focus on 99 original research articles. On the basis of the definitions established during the coding process, the research types were categorized into three groups: quantitative research, qualitative research, and mixed-methods research. The statistical results are presented in Fig. 5. As shown in Fig. 5, mixed-methods research constituted the largest proportion of existing studies in this field, with 48 articles. Quantitative research accounted for 43 articles, while qualitative research accounted for the fewest articles, at only 8 articles. This finding indicates that research in the field of technology and language learning anxiety predominantly employs mixed-methods and quantitative approaches. Importantly, the classification of quantitative, qualitative, and mixed-methods research can overlap to some extent. Consequently, this research employs the conceptual framework of mixed-methods studies articulated by Johnson, Onwuegbuzie, and Turner (2007), emphasizing that a study qualifies as mixed-methods exclusively when the quantitative and qualitative elements are present in approximately equal measures.Fig. 5: Categorization of research types and analytical methodologies.The classification highlights methodological diversity in technology-anxiety research, with mixed methods being most prevalent.Full size imageThe study also conducted a statistical analysis of the specific research methods employed in quantitative research, qualitative research, and mixed-methods research (99 articles in total). The results are presented in Fig. 5. Note that the “questionnaire method” refers exclusively to questionnaire surveys, and methods based on questionnaires, such as structural equation modeling, are not included in this category to avoid duplication.The methodological analysis of 99 original research articles reveals important insights into how FLA is being studied in technology-enhanced language learning contexts. The predominance of mixed methods approaches (48 articles) suggests that researchers recognize the need to combine quantitative measures of anxiety with qualitative exploration of learners’ subjective experiences. This approach is particularly relevant for FLA research, as anxiety manifests through both measurable physiological and behavioral indicators and personal, contextualized experiences. The substantial use of quantitative methods (43 articles), especially quasi-experimental (30%) and experimental methods (26%), indicates a strong focus on establishing causal relationships between technological interventions and anxiety reduction, which is crucial for evidence-based practice in FLA management. However, the limited number of purely qualitative studies (8 articles) points to a potential gap in deep, contextual understanding of how learners emotionally experience and cope with FLA in different technological environments.The specific methodological preferences are telling: The popularity of combined experimental and interview approaches (33% of mixed methods) suggests that researchers value both outcome measures and process understanding in FLA studies; the focus on controlled experiments in quantitative research reflects the field’s emphasis on isolating technology’s anxiety-alleviating effects; and the equal use of interviews and case studies in qualitative work indicates efforts to balance breadth and depth in understanding FLA phenomena. These methodological trends collectively suggest that while the field is developing robust ways to measure and intervene in technology-related FLA, there remains a need for more nuanced qualitative investigations into the lived experience of anxiety across different technological learning contexts. Future FLA research could benefit from expanded qualitative inquiry to capture contextual factors, more longitudinal mixed-methods designs to track anxiety trajectories, and methodological innovations in measuring the physiological aspects of anxiety during technology use.Grounded analysis for mechanisms of technology-related foreign language anxiety (RQ2)While “Empirical landscape of technology and foreign language anxiety research (RQ1)“ mapped the empirical landscape of existing studies, it remains unclear how specific technologies influence learners’ emotional responses. To address this, “Grounded analysis for mechanisms of technology-related foreign language anxiety (RQ2)“ employs grounded theory to uncover underlying mechanisms through which technology mitigates or exacerbates FLA.To address RQ2, this section synthesizes findings from grounded theory coding to delineate the dual mechanisms—alleviating and exacerbating—by which technology influences FLA. Many studies have deconstructed the inherent characteristics and attributes of technology from diverse perspectives, subsequently evaluating its positive and negative effects on anxiety and further analyzing the mechanisms underlying these contrasting outcomes. While there are overlapping aspects in the discussions regarding the positive and negative impacts of technology on language learning anxiety, no existing studies have systematically integrated these discussions.Discussions on the positive and negative impacts of technology on language learning anxiety are predominantly based on qualitative textual data, which are characterized by diverse presentation methods and varied perspectives. Because quantitative statistical methods are difficult to apply effectively in this context, this study adopts a grounded theory approach for analysis. Grounded theory is a theory-generating method that involves inductive analysis of qualitative data to identify themes or categories and is widely applied in the social sciences (Charmaz & Thornberg, 2021; Glaser & Strauss, 2017). Its primary analytical methods include concept extraction and coding analysis, which are used to distill key concepts and conduct in-depth data analysis from textual materials. In our study, following the grounded theory approach and utilizing NVivo 11.0 software, the included literature was analyzed to identify and code discussions related to technology’s impact on FLA, with a focus on risk perception and core advantages.Open coding: Identifying core concepts of anxiety-alleviating and anxiety-exacerbating mechanismsOpen coding, as a primary coding process in grounded theory, requires an open attitude, analyzing and coding the material line by line. Researchers must thoroughly explore the data to extract as many concepts as possible (Glaser & Strauss, 2017). Additionally, initial concepts should ideally be named using the participants’ own words to maintain accuracy and authenticity. During the coding process, it is also essential to compare and merge similar or repetitive concepts until the coding reaches saturation, meaning that no new concepts emerge. This approach ensures the extraction of as much information as possible from the interview materials and facilitates effective organization and analysis of the data.Regarding the reduction of anxiety through technology, on the basis of the coding analysis, we ultimately extracted 26 initial concepts from the literature, including “supportive, nonjudgmental space for speaking practice,” “absence of exposure to public criticism,” and “immediate feedback lowers anxiety.” Building on these initial concepts, we further summarized 17 initial categories, encompassing elements such as “psychological Safety,” “cognitive load management,” “enhanced self-efficacy,” “translanguaging Space,” “authentic and engaging content input,” and “immediacy of task feedback.” The specific classifications are shown in Table 3. These categories collectively constitute the positive effects of technology on language learning anxiety as identified in existing research.Table 3 Open coding of mechanisms by which technology alleviates FLA (partly).Full size tableDiscussions on the reduction of language learning anxiety through technology inevitably also lead to scholars’ attention to its potential to increase anxiety. Following the extraction and preliminary coding of the dimensions of potential risks in the included literature, 35 initial concepts were ultimately identified, including “excessive stimuli can exacerbate anxiety,” “lack of instant feedback,” and “induced anxiety by mimicking IELTS speaking examinations.” Building on these concepts, we further refined 18 initial categories, such as “high cognitive load,” “high-fidelity test simulations pressure,” “delayed feedback,” “dynamic difficulty adjustment,” and “lack of nonverbal cues.” As shown in Table 4, these 18 categories represent the aspects of technology that increase increasing language learning anxiety.Table 4 Open coding of mechanisms by which technology exacerbates FLA (partly).Full size tableAxial coding: Constructing thematic categories of technological impact on foreign language anxietyDuring the second coding phase, specifically the relational coding stage, an analysis of the initial concepts derived from the first coding is needed. The hierarchical relationships between primary and secondary categories are established by categorizing and understanding the internal relationships and logical sequences among the coded statements in the original material. Through relational analysis, researchers can deeply explore the connections among various codes and better grasp the overall research framework (Glaser & Strauss, 2017).We analyzed and summarized the 17 initial core strength categories and 19 initial potential risk categories obtained in the open coding phase in two primary category tables. The primary categories of core strengths and their connotations are shown in Table 5. Through further generalization of the initial categories, five primary categories were derived: psychological safety and emotional support, cognitive load and task management, feedback mechanisms and error tolerance, social interaction and peer dynamics, and autonomy and self-efficacy. The relational connotations of these primary categories, summarized from different perspectives, elucidate the positive impacts of technology on language learning anxiety, thereby laying a foundation for better understanding the current direction of research on the theme of technology and language learning anxiety.Table 5 Axial coding of mechanisms by which technology alleviates FLA (partly).Full size tableThe primary concerns and connotations of the initial categories of potential risk are detailed in Table 6. After summarization, five primary categories were identified: cognitive and task-related stressors, social and emotional disconnect, feedback and evaluation challenges, and technical and functional limitations. These primary categories are not isolated concepts; rather, their relationships and connotations, synthesized from multiple perspectives, comprehensively elucidate the mechanisms through which technology induces anxiety in language learning. This level of analysis enables a deeper exploration of the mechanisms by which technology affects language learning anxiety, providing robust support and guidance for future efforts to leverage technology to mitigate language learning anxiety.Table 6 Axial coding of mechanisms by which technology exacerbates FLA (partly).Full size tableSelective coding: Synthesizing mechanisms across technological dimensions and contextsOn the basis of the coding results of the impact of technology on language learning anxiety using grounded theory, we further developed the internal mechanisms and laid a logical framework of the influence of technology on FLA, as illustrated in Fig. 6.Fig. 6: The internal mechanisms and logical framework underlying technology’s impact on FLA.The conceptual model systematically classifies positive/negative effects, revealing technology’s paradoxical role in language learning anxiety management.Full size imageThe influence of technology on foreign language learning anxiety can be examined through two pathways—positive effects and negative effects—and four dimensions: the technological mechanism level, attribute level, teaching model level, and application scenario level. This framework elucidates the dual effects of technology on learners’ anxiety. From the perspective of technical attributes, adaptive learning systems provide personalized feedback, fostering learner confidence. However, the lack of emotional support in automated interactions may isolate learners, particularly when addressing complex linguistic challenges. In terms of the teaching model level, technology-driven personalized instruction optimizes learning outcomes by dynamically adjusting content. However, overreliance on technology may diminish the role of instructors, reducing emotional support and increasing anxiety. Additionally, technology-intensive models may demand higher levels of self-regulation, potentially overwhelming some learners. At the application scenario level, immersive technologies such as VR and AR create authentic language environments, thereby reducing anxiety by enhancing practical application. Nevertheless, the complexity and instability of technical devices may lead to frustration, especially during initial exposure or technical difficulties.In conclusion, leveraging technology in language learning requires a balanced approach. While the intelligent, personalized, and real-time feedback capabilities of technology should be fully utilized to increase learning efficiency, careful instructional design and application strategies must mitigate potential drawbacks, such as emotional detachment or technical challenges. This balanced integration is essential for advancing language education in alignment with contemporary technological advancements.Explaining the dual role of technology in foreign language anxiety: Theoretical integration and contradictionsThe apparent contradictions in research findings regarding the impact of technology on FLA can be systematically explained through multiple theoretical lenses that reveal the inherent complexity of this relationship. First, from a cognitive load perspective (Sweller, 2011), technological features inherently possess a “dual-edged” nature—the same functionality may serve as either beneficial germane load (enhancing learning) or detrimental extraneous load (impeding learning), depending on contextual implementation. This perspective explains why studies report opposing effects for similar technologies.Second, the technology acceptance model highlights crucial moderating variables. As identified in our coding results, individual differences in perceived usefulness and ease of use significantly influence whether learners experience a particular technology as anxiety-reducing or anxiety-inducing. For example, automated feedback systems that some learners find supportive (Zhang et al., 2024), others perceive as stressful (Hurd, 2007) —a conflict that reflects core TAM dimensions.Third, the multidimensional nature of language anxiety itself creates inherent contradictions. Technologies may simultaneously alleviate evaluation anxiety through private practice opportunities while exacerbating communication anxiety by removing nonverbal cues. This contradiction explains the polarized findings in studies examining the same technological intervention. Our grounded theory analysis of 99 studies naturally revealed these opposing categories through open coding, with “immediate feedback” (anxiety reducing) and “feedback delay” (anxiety inducing) emerging as two poles of the same dimension.These contradictions are not methodological flaws but rather reflect the authentic, context-dependent nature of technology-mediated language learning. Building on this understanding, our study provides a unified framework that not only explains these paradoxical findings but also offers practical guidance for tailoring technology-mediated language instruction to minimize anxiety while maximizing learning outcomes. Our grounded theory model thus represents a significant theoretical and practical advance over prior dual-role conceptualizations.Future agenda (RQ3)The dual-pathway framework underscores the complexity of technology’s emotional impact. Building on these insights, “Future agenda (RQ3)” proposes a future research agenda and stakeholder-specific strategies to translate findings into practice. In response to RQ3, this section articulates a future research agenda and provides targeted recommendations for key stakeholders. Presenting valuable future agendas is a crucial mission of systematic literature reviews, as evidenced by past practices (Prikshat et al., 2023; Pentina et al., 2023; Ancillai et al., 2023; Mariani et al., 2023). It is structured around theoretical development, empirical expansion, methodological innovation, and practical application. By identifying underexplored research topics in the field and highlighting gaps and weaknesses in current studies, systematic reviews significantly contribute to the future development of research areas.Recommendations for researchers: Theoretical, empirical, and methodological prioritiesOur literature analysis indicated that the interest in the impact of technology on language learning anxiety has increased. At the same time, the very low maturity of the field needs to be addressed. On the basis of the lessons learned, the following research agenda is proposed, as shown in Fig. 7.Fig. 7: Future agenda for researchers.The figure presents potential research pathways, covering technological, methodological, and theoretical dimensions.Full size imageFirst, the underdeveloped theoretical frameworks that currently hinder research on the impact of technology on language learning anxiety should be addressed. Specialized studies focusing on this specific area are scarce, and the literature lacks sufficient depth for full explorations of the complexities of anxiety in technology-mediated language learning environments. To bridge this gap, more scholarly efforts are needed to develop a comprehensive taxonomy of anxiety types and influencing factors in technology-enhanced language learning contexts. Additionally, creating a unified theoretical framework that integrates insights from language learning anxiety and technology-mediated education could provide a stronger foundation for future research. By prioritizing these areas, researchers can uncover valuable insights into how technology shapes anxiety in language learners, ultimately leading to the development of more effective strategies to mitigate anxiety and enhance learning outcomes.Second, future research should prioritize addressing three critical gaps in the study of the impact of technology on language learning anxiety: first, by expanding investigations beyond the predominant focus on higher education to include other educational stages, such as primary, secondary; second, by exploring the often overlooked areas of reading, writing, listening and other fields to gain a more comprehensive understanding of how anxiety manifests across the full spectrum of language skills; and third, by leveraging the potential of virtual reality VR and AR to uncover the mechanisms of technology-induced anxiety, despite the current challenges of high costs and technical complexities (Radianti et al., 2020). While speaking tasks have dominated existing studies, the emotional and cognitive challenges associated with technology use in reading, writing, listening, and other fields remain understudied, which limits our understanding of how anxiety operates in these critical areas. Similarly, the majority of research has been conducted within higher education settings, leaving the experiences of learners in other educational stages underexplored. Furthermore, the integration of VR and AR into this line of inquiry could offer innovative ways to simulate immersive learning environments and uncover the specific triggers of technology-induced anxiety, although their implementation may require addressing significant barriers (Wang et al., 2025d). By adopting a more holistic approach, expanding the scope of research to include diverse educational stages and language skills, and exploring the potential of emerging technologies, future studies can contribute to a more nuanced understanding of technology-related language learning anxiety and inform the development of evidence-based strategies to support learners across all educational levels.Third, to better understand the impact of technology on language learning anxiety and to derive actionable insights, more rigorous evaluation methods are needed. Current research predominantly relies on self-assessment questionnaires to measure students’ anxiety levels, which, while useful, lack the precision and scientific robustness that can be achieved through the use of physiological measurement tools, such as electroencephalography or other neuroscientific instruments (Kelsen et al., 2025). These tools can provide objective and accurate data on stress and anxiety levels, thus offering a more comprehensive understanding of the emotional and cognitive states of language learners in technology-mediated environments. Future studies should therefore integrate both quantitative and qualitative research methods to assess not only learners’ self-reported experiences but also their physiological responses. This dual approach would enable researchers to identify patterns and correlations that might otherwise remain undetected through self-assessment alone. Additionally, evaluations of technology-enhanced language learning applications should be conducted from both technical and pedagogical perspectives to ensure that they meet the needs of both teachers and students. To achieve this goal, future research could incorporate workshops, surveys, and focus group discussions to gather insights from educators and learners alike, ultimately informing the design of more effective and anxiety-reducing language learning technologies. By adopting such a multifaceted approach, researchers can develop a deeper understanding of the relationship between technology use and language learning anxiety, paving the way for the creation of evidence-based interventions and tools.Recommendations for lecturers: Integrating learner-centered principles with technology to reduce language learning anxietyBuilding upon grounded theory analysis and learner-centered educational principles, this study proposes four evidence-based recommendations for lecturers to mitigate foreign language anxiety through technology integration while promoting active language learning. These recommendations align with the paradigm shift from traditional teacher-centered instruction to approaches that prioritize student engagement, agency, and psychological safety (McCombs & Whisler, 1997).First, lecturers should implement cognitive load management strategies through carefully selected active learning technologies. Research has demonstrated that interactive tools such as subtitled videos (Chen, 2011) and structured digital environments can reduce cognitive strain while promoting language production when aligned with learner-centered principles of gradual skill development. This approach supports the emphasis of learner-centered educational principles on developmental appropriateness while encouraging meaningful language practice through technology-enhanced tasks that allow student input in activity selection.Second, feedback systems should combine technological efficiency with opportunities for active language use. While immediate automated feedback can reduce uncertainty (Harb et al., 2014), lecturers should combine this tool with reflective dialog activities that foster both self-efficacy and language output (McCombs & Whisler, 1997). This dual approach addresses cognitive needs through timely corrections while creating authentic contexts for language practice, mirroring findings that active, learner-centered classrooms yield greater motivation and linguistic achievement.Third, psychological safety should be prioritized through technologies that enable active yet low-anxiety language practices. Tools such as virtual worlds (Melchor-Couto, 2017) and AI avatars can reduce evaluation apprehension while providing opportunities for meaningful interaction within a learner-centered framework that encourages questioning and risk-taking. Online collaborative spaces (Ai et al., 2024) further support the principle of learner-centered education in language learning through technology-mediated communicative activities.Finally, lecturers should cultivate shared responsibility for learning through technology-mediated active learning choices. Allowing students to select from vetted digital tools for language practice (e.g., choosing between VR simulations or discussion forums) operationalizes the learner-centered tenet of student-directed learning. Research confirms that such autonomy-supportive practices, when combined with appropriate technological scaffolding, enhance both linguistic competence and comfort in active language use.Recommendations for learners: Integrating positive psychology with technology-driven language learningModern language learners can effectively overcome FLA by strategically combining technological tools with positive psychological approaches. These evidence-based approaches create optimal conditions for language acquisition by addressing both the cognitive and the affective dimensions of learning.First, learners should cultivate self-efficacy—the belief in their ability to succeed—to reduce anxiety and build confidence (Hong et al., 2019). Grounded in the emphasis of positive psychology on strengths (Seligman & Csikszentmihalyi, 2000), learners can achieve this by setting incremental goals (e.g., mastering five new words daily) and celebrating small wins, which reinforce competence. Collaborative technologies such as multiplayer language games enhance collective self-efficacy through shared achievement (Hong et al., 2019), while reflective journaling (Oxford, 2016a) fosters mastery. Emotional intelligence (EI) training—such as identifying stressors in digital tasks (Barchard et al., 2016)—can further buffer anxiety, as EI correlates with lower FLA (Dewaele et al., 2008).Second, translanguaging and personalized learning can be synergized to create low-anxiety environments. Translanguaging (e.g., using AI assistants that permit language-switching; Zhang et al., 2024) aligns with the focus of the EMPATHICS model on emotional safety (Oxford, 2016b), reducing fear of judgment. This approach can be paired with adaptive tools (e.g., AI platforms that adjust difficulty; Yang et al., 2020) to tailor pacing, a key factor in anxiety reduction (Hurd, 2007). For example, replaying VR dialog simulations leverages embodied cognition while allowing self-paced repetition, addressing both cognitive and emotional needs.Third, active learning strategies should be integrated with technological affordances. AI can be used for real-time feedback on pronunciation to target specific anxieties (Pishghadam, 2009), and online communities can be used to practice EI-driven communication (Gkonou & Mercer, 2017). Streaks on platforms such as Duolingo tap into a growth mindset, while mindfulness apps can preempt technostress (Dewaele & MacIntyre, 2024). By merging positive psychology principles (e.g., flourishing via strengths) with learner-centered tech, learners transform anxiety into engagement, echoing findings that EI mediates second language success (Shao et al., 2013).Recommendations for developers: designing emotionally responsive language learning technologiesTo reduce user anxiety in the development of language and educational technology, developers should prioritize user-centered design principles, incorporate effective feedback mechanisms, and foster a sense of psychological safety. These strategies are supported by empirical research and can significantly enhance the user experience by minimizing stress and promoting engagement.First, developers should adopt user-centered design principles to ensure that technological tools align with users’ cognitive and emotional needs. A high cognitive load, caused by complex interfaces or excessive information, is a major source of anxiety (Salem, 2019). To address this, developers should design intuitive and streamlined interfaces that reduce cognitive strain. For example, incorporating subtitles in videos or breaking down tasks into smaller, manageable steps can increase comprehension and reduce anxiety (Chen, 2011). Additionally, developers should avoid overwhelming users with multimodal information overload, as simultaneous content delivery can lead to confusion and stress (İnceçay & Koçoğlu, 2017). By prioritizing simplicity and clarity in design, developers can create tools that are accessible and less anxiety-provoking.Second, effective feedback and “learning as assessment” mechanisms should be integrated into educational technologies to provide users with timely, accurate, and constructive feedback (Liu et al., 2024b). Delayed or ambiguous feedback can heighten anxiety by leaving users uncertain about their progress (Hurd, 2007). Developers should implement systems that offer immediate and interactive feedback, which has been shown to lower anxiety and enhance learning outcomes (Harb et al., 2014). However, automated feedback tools must be carefully designed to avoid providing confusing or redundant responses that can increase frustration (Sari & Han, 2024). Personalized feedback that addresses individual learning needs can further reduce anxiety by making users feel supported and understood. By ensuring that feedback is both timely and meaningful, developers can help users build confidence and reduce stress.Third, fostering psychological safety should be a core objective in the design of educational technologies. Anonymity and invisibility features, such as those found in virtual worlds or audiographic conferences, can significantly reduce the fear of negative evaluation by allowing users to practice without judgment (Melchor-Couto, 2017). Developers should also incorporate emotionally supportive elements, such as human-like avatars or AI-driven agents that provide positive reinforcement, to increase users’ emotional well-being and reduce pressure (Wang et al., 2024a). Furthermore, creating opportunities for social interaction and peer support, such as online forums or collaborative platforms, can alleviate feelings of isolation and build a sense of community (Ai et al., 2024). By designing tools that promote psychological safety, developers can create an environment where users feel comfortable taking risks and expressing themselves. The comprehensive representation of the forthcoming agenda is presented in Fig. 8.Fig. 8: Strategic framework for future directions across four key dimensions.The model points out future research directions and bridges pedagogical, technological, and psychological dimensions to address anxiety holistically.Full size imageConclusion, implication, and prospectThis final section synthesizes the study’s main findings, theoretical contributions, and practical significance, reflecting on its implications for the broader field.ConclusionOur study addresses the three initial research questions through a systematic review of 99 articles on the impact of technology on language anxiety.First, concerning the core dimensions of the research framework (RQ1), the limited proportion of literature supported by theoretical frameworks suggests the field’s immaturity and significant potential for growth. The scarcity of studies focused exclusively on anxiety may reflect the pervasive and complex nature of this phenomenon, which often coexists with other psychological or physiological issues. Measurement methods predominantly rely on self-assessment tools such as questionnaires and interviews; future research could incorporate neuroscientific approaches such as electroencephalography to increase objectivity and effectiveness. The minimal use of AR&VR in these studies highlights substantial opportunities for leveraging such technologies to reduce language learner anxiety. Additionally, the analyzed literature focuses primarily on the impact of technology on speaking anxiety, with less attention given to other language skills, and predominantly focuses on higher education contexts. Furthermore, in terms of mechanisms (RQ2), the study employs grounded theory to deconstruct how technology exacerbates or alleviates foreign language learning anxiety, proposing application models that capitalize on strengths and mitigate weaknesses. Finally, on the basis of a review of the literature, this study identifies several promising research directions from four perspectives for future analysis (RQ3).Our literature analysis reveals a growing interest in the impact of technology on language learning anxiety, yet the field remains underdeveloped. To address this issue, we propose a comprehensive research agenda identifying three critical research gaps and delineating four primary research directions. Furthermore, we offer targeted recommendations from three distinct perspectives—those pertaining to lecturers, learners, and engineering developers. This multifaceted approach aims to provide a robust framework for advancing understanding and practice in the field.Limitations and future researchOwing to the nature of the search and selection process, our work has several limitations. First, we focused only on the keywords related to technology, ChatGPT, and AI etc. However, the field of technology is vast and diverse, and there may be other relevant terms, tools, or innovations that were not included in our search criteria. For example, emerging technologies such as natural language processing, machine learning algorithms, or specific applications such as robotics, autonomous systems, or the Internet of Things might have been overlooked. These areas could have provided additional insights or literature that are highly relevant to our study. Furthermore, the rapid evolution of technology means that new tools, frameworks, or research related to AI and ChatGPT might have been published after our search period, which could further enrich our analysis. Importantly, the narrow scope of our keywords may have resulted in a less comprehensive collection of literature. Future studies could benefit from expanding the search terms and exploring a broader range of technological advancements to provide a more holistic understanding of the topic.While our systematic review methodology provides a comprehensive analysis of the impact of technology on language learning anxiety, several important limitations must be acknowledged. The study’s focus on specific databases and finite sample size may have excluded relevant perspectives, and researcher subjectivity in data interpretation could introduce bias. However, our analysis also reveals crucial insights into positive psychological factors that can mitigate technostress in technology-assisted language learning. Notably, learner-centered approaches have emerged as particularly effective in reducing anxiety. Adaptive learning technologies that personalize content and self-paced platforms empower learners, fostering autonomy and control, key factors in anxiety reduction. Furthermore, active learning strategies facilitated by digital tools, such as gamified applications and VR simulations, create engaging, low-pressure environments that promote language practice while minimizing performance anxiety.These findings suggest that future research should not only address the current limitations through expanded database coverage and larger samples, but should also specifically investigate: how different technologies support learner autonomy and self-regulation, the role of positive emotional design in educational technology, and strategies for balancing technological challenges with psychological support mechanisms (Wang et al., 2025e). Such investigations would provide a more holistic understanding of how to harness technology’s potential while safeguarding learners’ psychological well-being in foreign language acquisition.