The adoption of AI tools in English language teaching has dramatically increase in recent years causing a

fundamental shift to lesson planning and teaching at all levels of countries. The purpose of this study is to

investigate and understand how Malaysian primary English teachers employ AI in designing and implementing

lesson plans for students and describe their perceptions of the impact of AI on students. This study responds to a

critical research gap by investigating the underexplored experiences of primary level teachers to whom learners’

expectations for engaging, scaffolded and contextualised learning are not the same as those of secondary and

tertiary educators. Using a qualitative approach, data were gathered from six experienced primary teachers in

different Malaysian schools through semi-structured interviews, observations and analysis of AI enhanced lesson

the reliability of AI generated content and the risk of relying too much on them. In conclusion, teacher agency,

prompt engineering and reflective practice are the keys to the effective integration of AI in primary English

classroom.

The rapid and widespread adoption of Artificial Intelligence (AI) tools, particularly sophisticated generative

language models, marks a crucial turning point in global education (Sallam, 2024; UNESCO, 2023). These

technologies are increasingly moving beyond administrative support to fundamentally reshape pedagogical core

practices, most notably teacher lesson planning (Gleason, 2022). For the field of English Language Teaching

(ELT), lesson planning is the critical point where instructional theory translates into actionable design, dictating

to complex classroom realities (Hwang & Tu, 2021). This shift challenges the traditional definition of teacher

expertise and raises crucial questions about the genuine impact on student outcomes.

While some studies broadly note AI's use in lesson preparation, there is a pronounced scarcity of research that

qualitatively explores the specific, detailed process and rationale by which primary teachers utilise AI to create

concrete instructional resources such as tailored games, differentiated prompts or customised feedback scripts

within their daily workflow (Kerr and Kim, 2024). Besides, the majority of qualitative findings on teacher AI use

are derived from secondary or higher education settings (Lan and Zhou, 2025). The experiences of Primary

English teachers who prioritise foundational literacy, scaffolding and highly engaging activities for young

learners are distinctly underexplored. Their unique pedagogical needs with respect to AI warrant specific

investigation. Furthermore, although efficiency gains are often cited, the literature lacks rich, descriptive

The research objectives of this study are:

English language learning.

This study will focus on 3 research questions, which are:

instructional materials during their lesson planning?

A rich sources in recent literature tells the tale of English teachers and their use of AI to develop more efficient

and effective lessons. Across contexts, the use of AI tools is said to replace human workers or support three

interconnected functions such as the generation and organising of learning resources, alignment with standards

and rubrics for assessment as well as informing data planning in order to customise instruction according to

for discussion, rubrics for writing and the sequences of activities which the teachers have to critically take up,

adapt and contextualise in relation to their classrooms. Belloula (2024) and Wang and Li (2024) reported that

AI-generated products have the potential to enhance or expand rather than limit the instructional aggregation.

instruction that takes into account heterogeneous levels of proficiency (Belloula, 2024; Roslan & Mohamod,

2025). The potential AI in this context relies heavily on teacher skills which include prompting practices, data

literacy and critical organising of machine-generated output. Therefore, Tan et al. (2025) and Xiaofan (2025)

Access and equity issues surface time and with fears about the digital divide defining potential over-dependence

on algorithmic decision making in schools that have no governance structures promoting reflective practice and

ethical use (Vesna et al., 2025). As a whole, AI is presented as a productive co-planner that can provide planning

help with professional knowledge as well as alignment with standards and sustained teacher support for prompts

and assessment of responses (Dogan, 2023; Zhang & Tur, 2024).

confidence particularly if feedback is aligned with the targeted language norms and embedded in communicative

tasks (Wei, 2023; Zou et al., 2025).

Moorhouse et al. (2024) and Granström and Oppi (2025) also reported that in reading and listening, AI tools

populate texts based on learner interests and ability which have been correlated to increased engagement and

emergent comprehension gains. In all of these areas, an emerging agreement is that the effectiveness of AI

depends on teacher organising, task design and explicit scaffolding. In other way of saying, teachers serve as

organisers and coaches who mediate AI-generated content by guiding students through it to achieve learning

goals and standards. Ma and Chen (2025) mentioned that teachers also provide language learners with the human

relational context necessary for them to engage with that content. Prior reviews indicate that English teachers

development and governance which maintains teacher control over tasks while promoting ethical use (Granström

et al., 2025; Wang et al., 2024).

Central goals of AI-mediated ELT are to improve Language Learning and Student Engagement (HESTME)

wherein it has been reported that "personalised, interactive AI experiences increase motivation, persistence and

positive attitudes toward learning English when feedback is timely and tasks relate to real communicative needs".

The language efficacy returns the greatest improvements in writing quality and speaking proficiency when AI

feedback is accurate, supportive and embedded into a scaffolded pedagogical sequence that reflects assessment

criteria (Wei, 2023) to use as a guide.

Nevertheless, tension that comes from distraction, lack of face-to-face interaction snore quality and too much

allocation (Kim & Kim, 2022).

Ethical concerns about for example data privacy, transparency and bias have the potential to impact on

engagement and achievement, determining how systems are taken up, trusted and governed within classroom

practice that span the wide variety of context in which has been tested (Wang et al., 2024).

The Technological Pedagogical Content Knowledge (TPACK) framework, first conceptualized by Mishra and

Koehler (2006), provides a robust theoretical foundation for understanding how teachers integrate technology

Knowledge (PCK) by incorporating technology as a critical dimension of modern teaching practice. Within the

context of English language education, TPACK emphasises the dynamic interaction among technological

approaches were better able to design lessons that fostered engagement and comprehension, particularly when

using digital tools to enhance reading instruction. In the Malaysian context, Shafie et al. (2019) reported that

In relation to this study, the TPACK framework serves as a conceptual lens for examining how primary English

teachers describe their process of utilizing AI tools to generate and customise instructional materials during lesson

technological affordances of AI such as adaptive worksheets or AI-generated vocabulary games reflects the

central tenets of TPACK. By analysing how teachers navigate this interplay, the study seeks to understand the

learners actively construct knowledge through social interaction, experience, and reflection rather than passively

meaning-making, encouraging activities that promote authentic communication, collaboration, and problem-

solving (Aljohani, 2017). Within classrooms where AI tools are integrated, constructivist principles are

manifested when technology is used to facilitate interactive, student centered experiences such as AI chat

simulations, grammar games or adaptive storytelling platforms that prompt learners to co-construct meaning

dynamically.

Empirical studies have confirmed the strong link between constructivism and effective technology integration.

Zhang and Miao (2025) highlighted that constructivist learning environments foster deeper language acquisition

by encouraging exploration and contextualised use of language particularly when supported by digital tools.

Similarly, Liu et al. (2024) emphasised that integrating AI and digital technologies into English language teaching

allows teachers to create more meaningful learning contexts that encourage collaboration and self-regulated

views teaching as a process of continual learning and adaptation, in which teachers act as reflective practitioners

They are engaging in the kind of reflective practice that constructivist theory endorses as essential for professional

growth.

In sum, both TPACK and Constructivist Learning Theory underpin this study’s exploration of AI integration in

English language teaching. TPACK informs the structural and cognitive dimensions of how teachers plan and

deliver AI assisted lessons while constructivism provides insight into how these practices affect student learning

and teacher professional identity within the evolving landscape of AI enhanced education.

Following this, the participants provided video recordings of the two corresponding lessons that were conducted

based on the submitted lesson plans. The researcher reviewed the recorded lessons and prepared fieldnotes using

the Classroom Observation Form. To ensure accuracy and credibility, the fieldnotes were returned to the

participants for member checking, allowing them to verify that the observational notes accurately reflect what

occurred in the recorded lessons.