AI tools significantly enhance K-12 students' English language learning experience, with chatbots and AI personal assistants particularly effective for developing speaking skills in EFL contexts.

Objective: This narrative research synthesis aimed to investigate the use of AI tools to support English language learning in K-12 ESL/EFL classrooms by examining what AI tools are currently used, how they support specific language skills, the perceptions of teachers and students about their effectiveness, challenges, and limitations, and to identify important topics and recommendations from existing research.

Methods: The researcher conducted a systematic electronic search using the databases ERIC and PsycINFO to identify peer-reviewed academic journal articles published between 2022 and 2024 about AI tools used in ESL/EFL K-12 classrooms. After applying inclusion and exclusion criteria, 23 empirical studies were selected for analysis. A deductive coding approach based on five research questions was used to analyze the data, followed by qualitative inductive analysis to identify themes and patterns. The study was informed by Second Language Acquisition theories of input, output, and interaction (Krashen, Swain, and Long, 1985) and the taxonomy of AI tools in education (Holmes and Tuomi, 2022).

Key Findings:

• AI chatbots (43.5%) and AI personal assistants (17.4%) were the most commonly used AI tools in ESL/EFL K-12 classrooms, followed by other AI tools like speech recognition, image recognition, and natural language processing (39.1%).
• The vast majority of studies (21 out of 23) were conducted in Asian countries, particularly Taiwan (8), China (6), South Korea (5), and Hong Kong (2), with very little research from ESL contexts.
• Speaking skills were the most frequently supported language skill (15 studies), followed by reading (10 studies), writing (8 studies), and listening (7 studies).
• Vocabulary acquisition was a central focus across multiple skills development, playing a crucial role in connecting different domains and supporting multiple AI-driven tools.
• AI tools helped reduce student anxiety in language learning contexts and provided personalized, immediate feedback that traditional classroom settings often couldn't offer.
• Teachers and students generally had positive perceptions of AI tools' effectiveness, although they identified challenges such as technical limitations, lack of institutional support, and privacy concerns.
• English proficiency level significantly influenced how students interacted with and benefited from AI tools, with higher-proficiency students perceiving more opportunities while lower-proficiency students experienced more limitations.
• AI tools facilitated a more student-centered approach both inside and outside the classroom, supporting autonomous learning and addressing specific student needs.
• Multimodal AI tools combining text, images, and audio were particularly effective for vocabulary learning and engagement.

Implications: This research synthesis offers valuable insights into the application of AI tools in K-12 language education, demonstrating their potential to transform traditional language learning environments. The findings suggest that AI tools can effectively supplement teacher instruction by providing personalized feedback, reducing language anxiety, supporting autonomous learning, and addressing the limitations of traditional classrooms where students have limited opportunities to practice speaking skills. The research also highlights the importance of considering student proficiency levels and providing appropriate teacher support when implementing AI tools, as their effectiveness varies based on these factors. By understanding both the benefits and challenges of AI integration, educators can make more informed decisions about how to effectively incorporate these tools into language instruction, ultimately creating more engaging and effective learning experiences for K-12 English language learners.

Limitations: The study acknowledges several limitations. First, the research primarily reflects AI tool use in EFL contexts in Asian countries, with minimal representation from ESL contexts or Western educational settings. Second, the three-year timeframe (2022-2024) is relatively short, though justified by the rapid evolution of AI tools during this period. Third, most studies involved small sample sizes and short-term interventions, making it difficult to assess long-term impact or generalizability. Finally, many studies focused on the novelty effect of AI tools rather than their sustained pedagogical value, and few addressed important concerns about data privacy, accessibility, and academic integrity in depth.

Future Directions: The research identifies several promising directions for future study:

• More diverse research contexts are needed, particularly in ESL settings and non-Asian countries, to understand how cultural and educational differences may influence AI tool effectiveness.
• Larger-scale studies with longer-term implementations would provide more reliable evidence about AI tools' sustained impact on language learning.
• More in-depth exploration of safety, privacy, and academic integrity issues is necessary, especially considering the vulnerable K-12 population.
• Research on teacher training and support systems for AI implementation would address a critical gap in current knowledge.
• Future studies should examine how to customize AI tools for different proficiency levels and learning needs to maximize their effectiveness.
• Additional research on multimodal applications of AI in language learning could further enhance understanding of how these tools support comprehensive language development.

Title and Authors: "AI Tools to Support English Language Learning in K-12 Classrooms" by Mauricio Antonio Dos Santos

Published On: 2025

Published By: The Ohio State University (Master's Thesis)