Teaching of English language in Bangladesh is undergoing a metamorphosis because of the advent of ICT and the use of AI. In light of research on the use of ICTs and materials in public tertiary level education in the context of EFL pedagogy, this paper explores the possible impacts of AIbased ICT tools on the quality EFL tertiary level classroom instruction. This background highlights the established status of Bangladeshi graduates' communicative proficiency despite the uninterrupted presence of English. The study employs a mixedmethods approach, including surveys and interviews, to investigate the impact of AI tools on learner motivation, autonomy, and language skill development. The findings indicate that AI in the education sector is useful for boosting learner engagement, facilitating a customized learning experience, and shortening the educational gap. But obstacles remain in infrastructure, teacher preparedness and reliance on the outputs of A.I. Students cited increased confidence in speaking and writing, while teachers considered AI an addon rather than a replacement for their resource library. Recommendations are organized along lines of systematic teacher education, judicious AI implementation, and stronger digital literacy components in ELT programs. The study concludes arguing that as ICT is insufficient to enhance language proficiency and testing (LPT) and communicative competence (CC) of EFL learners in Bangladesh alone. Nevertheless, it can with the use of AIassisted ICT in ELT brings a revolution and a new dimension of attaining LPT and CC.
There have been cyclical changes in the field of ELT in Bangladesh over the years in response to global pedagogical trends and local needs. In the era of globalisation where good communication is the key to success, graduates of Bangladesh have been encountering a challenge for ages seven years' learning of English results in prompted devastating as far as skills like speaking, writing, and critical thinking are concerned
University teachers have been growing in the last few years, as is the case of the infusion of Information and Communication Technologies (ICTs) in ELT. Research works reported that multimedia presentations, language learning apps, and digital libraries have the potential to enhance classroom interactivity and learner's motivation
AI in education isn't some faroff future-it's an increasingly present one. Smart apps such as Duolingo's intelligent algorithm, ELSA Speak's speech analytics engine, and AI chatbots like ChatGPT provide automated realtime feedback, error correction, and tailored learning pathways for learners
In Bangladesh, AI based ICT tools are increasingly being used in language education in elite and urban schools by educators
The theoretical framework is offered by the traditions of Technology, Communication, and Society and Education and Critical Discourse. It examines the role of technologymediated communication tools in not only facilitating second language learning but restructuring the learnerteacher relationship, reconceptualising feedback loops, and supporting the culture(s) of autonomous language learning
Building on early works concerning ICTrelated tools for teaching English in the context of Bangladesh, this paper draws attention to the development in tools from static to intelligent systems. It raises the question not just of what students can learn but of how they can do so-and of how AIdriven technology can advance equity, individualisation, and attention to audience in the communication of tomorrow
1. To what extent do AIbased ICT tools impact EFL learners' language skills (writing, speaking, listening, and reading) at the tertiary level in Bangladesh? 2. How do AIintegrated ICT platforms affect learner autonomy, engagement, and motivation in EFL classrooms? 3. What are the perceptions of EFL teachers and students regarding the benefits, challenges, and ethical considerations of using AI tools in English language instruction?
2. Literature Review Information and Communication Technology has been shaping the educational environment, giving language learners new instruments and teachers new educational approaches. In this international view, the introduction of ICT in English Language Teaching (ELT) means, at the same time, to move from a somewhat teacher centred and to a learnercentered education. ICT has not been equally transported by the public and private sectors in Bangladesh
More generally, ELT is 'moving from basic ICT tools - like PowerPoint and YouTube - to technologies that are based on AI (Artificial Intelligence) that will provide personalisation, automation, and even adaptation'
Autonomy Definition According to
Teachers are important gatekeepers when it comes to the use of AI tools. A group is eager to embrace AI, another is hesitant of it because of insufficient preparation and doubts related to pedagogical alignment
Automatic Writing Assessment (AWA) tools, for example Grammarly and Write & Improve, have revolutionized writing teaching. Qasem, M., & Viswanathan, K. (2022) discovered that students who used AWE frequently produced work that was more accurate and more organized. These are tools which provide grammar correction, vocabulary reminders, and style enhancement - and features that have always taken a lot of teacher hours for correction. Furthermore, some AWE tools are useful up to a point," and that the tools must be in combination with teacher feedback for deep learning
Speaking is still the most difficult skill to teach and to evaluate. Chatbots (bot) with AI technology are able to have conversation with people giving an opportunity to learn in low anxiety situations
ELT curricula, hence, should now account for AI as an emerging powerful force. British Council in 2021 states that it is vital to include AI literacy in language education in order to produce futureready graduates. KukulskaHulme,
AI has many benefits; however, it does not come without risks. Educational technology might perpetuate social disparities, if the access to it is unequal. This is still an issue in Bangladesh, especially for those rural or poor learners
Empirical AI in ELT research come from global north. There is limited literature in South Asian settings including in relation to collegelevel students in semiurban areas. This paper attempts to fill this gap by investigating the interactions of Bangladeshi learners and teachers with AI tools with culturally and pedagogically resonant insights. It might also be worthwhile to explore the longterm effects that AI might have on students' writing development, if chatbotbased speaking tasks really are more effective than traditional ones: and how traditional and AI integrated classes look like in comparison. There is also potential for cross disciplinary studies that compare AI use in ELT with social studies, sociology, linguistics, and educational policy.
Methodology This is a mixedmethods research, using both quantitative and qualitative methods to explore the dimension of AIbased ICT tools in ELT at some private universities in Bangladesh. While structured surveys provided the quantitative information and semistructured interviews were used for qualitative data.
The research was conducted between two public universities in Bangladesh; Jagannath University and Comilla University. The participants were 50 EFL learners studying general English courses and 10 EFL teachers. The researcher sampled students who use AIbased tools like Grammarly, Duolingo, ChatGPT, etc., in a purposive manner.
Two major instruments were designed: a) a structured questionnaire with both multiplechoice and Likerttype questions that students completed and assessed whether they had worked with AIbased ICT tools; and b) an interview guide for teachers that focused on their ICT practice and knowledge, their perception of AI use, its effectiveness, and challenges in its adoption. Online surveys were sent using Google Forms and interviews were done via Zoom with participant consent with recorded interviews.
For quantitative data, descriptive and inferential use of SPSS and MS Excel was made for (frequencies, Average, percentage). Interview transcripts were analysed using thematic analysis with a focus on emergent patterns such as learner autonomy, usability of AI, teacher attitudes, and infrastructural constraints. The triangulation of sources also strengthened the study and provided a comprehensive view of how AIsupported ICT tools are impacting the ELT of Bangladeshi private universities.
A notable response recorded from the survey participants that significantly impacts this study is as follows:
Along with the quantitative data, teachers' and students' openended answers shed more light on their opinions, difficulties, and experiences with the application of AI tools in ELT. Occasional Users 22%
High levels of technological engagement in ELT were demonstrated by the vast majority of students (78%) who reported regularly using AI apps like Grammarly, ChatGPT, and Duolingo. The 22% of those who used them occasionally probably had problems with training, access, or their preference for conventional learning methods. This disparity highlights the necessity of inclusive digital literacy initiatives. After utilising AI technologies, 75% of students in this portion (figure 3) said they felt more confident about their speaking and writing skills. These platforms offer instantaneous, confidential feedback, which probably lessens anxiety and promotes more confident communication. Individual learning variations or a lack of appropriate guidance on how to use such tools effectively are reflected in the modest percentage of disagreement. Poor 10%
AIgenerated feedback was rated as excellent or good by 70% of students, indicating that most of them felt it to be relevant and helpful. Nonetheless, 30% said the response was mediocre or subpar, indicating a lack of precision, customisation, or user comprehension. This demonstrates the necessity of integrating guided human training with AI tools. Seventy percent of students were inspired to use AI technologies for solo English practice. This implies that elements like autonomy, gamification, and immediate outcomes successfully encourage lifelong learning outside of the classroom. Nonetheless, 30% expressed no opinion or disagreement, suggesting a range of motivated reactions and the necessity of customizing tools to suit personal preferences.
15.00%
Remarkably, 85% of students were in favor of using AI tools in the classroom. This indicates that students have a favourable attitude toward blended learning approaches, which use technology to enhance inperson training. The 15% opposition highlights the necessity for structured AI adoption and may be a reflection of reluctance to change or bad experiences in the past.
20.00%
No (Not Concerned)
Concerning the ethical usage of AI tools by students, 80% of educators pointed out dangers such as plagiarism and blind dependence. In order to encourage responsible tool usage, teachers themselves expressed uncertainty regarding appropriate guidelines, highlighting the necessity of curriculumintegrated digital ethics training and AI literacy modules. Poor internet and infrastructure 8 2.
Lack of access to premium AI tools 7 3.
No teacher training on AI 9 5.
No institutional guidelines/support 6
The majority of teachers cited a lack of professional training and inadequate infrastructure as the two main obstacles to the use of AI. AI integration in ELT becomes erratic in the absence of institutional support, tool accessibility, and internet dependability. Instructors also bemoaned the lack of exposure to AIfocused capacity building initiatives.
By offering a relaxed setting for autonomous language practice and development, AI tools promoted numerous revisions, increased students' confidence in editing and selfcorrection, and decreased their speaking fear.
Students reported that they were able to identify specific writing and speaking shortcomings and return their work more frequently thanks to AI feedback. They were able to selfcorrect without waiting for teacher feedback thanks to tools like Grammarly, which increased their independence, enhanced their fluency, and lessened the stress associated with performing well on important assignments.
Teachers noted that by letting pupils study at their own pace, AI promotes individualized learning. Students, however, frequently rely too much on AIgenerated content without fully comprehending it. Concerns regarding ethical awareness, plagiarism hazards, and the necessity of responsible AI use standards are raised by the fact that teachers themselves lack the necessary skills to integrate AI properly.
Teachers concurred that tailored learning and increased classroom engagement were made possible by AI tools. However, issues with abuse, overreliance, and a lack of ethical awareness were brought up. Additionally, many felt unprepared to effectively incorporate AI into their teaching, indicating the need for institutional support and formal training.
Teachers highlighted dangers such as naïve reliance on AI outputs, shallow learning, and plagiarism. In order to guarantee the ethical, responsible, and pedagogically sound use of AI in ELT, many teachers also felt undertrained in digital pedagogy and demanded institutional measures.
Instructors emphasized that restricted access to AI tools and inadequate internet connectivity impede successful integration in the classroom. Additionally, they cited the lack of professional development opportunities and institutional support, which left them illequipped to effectively apply AI technology in English language instruction.
"We don't have stable internet to use AI tools regularly." "There are no workshops to help us use AI effectively." Teachers' dissatisfaction with the absence of professional assistance and the poor infrastructure is reflected in these statements. They can't use AI tools in the classroom regularly without a reliable internet connection. Furthermore, the lack of workshops or training limits the potential advantages of AI in language teaching and learning by preventing teachers from comprehending and successfully using it.
Instructors have noted that plagiarism and improper usage of AIgenerated content are frequently caused by students' ignorance of AI ethics. A lot of students turn in work without knowing what originality issues are. The necessity for explicit rules and ethical training is highlighted by the fact that teachers themselves are unsure about institutional policies regulating the use of AI.
"Students copypaste from AI without knowing it's unethical." "We weren't trained to teach students about responsible AI use." These results show that students frequently turn in AIgenerated content without being aware of the moral ramifications, such plagiarism or a lack of originality. Instructors acknowledge that they do not have the necessary formal training to teach students how to utilize AI responsibly. This reveals a serious ethical literacy gap that leaves teachers and students open to inappropriate, careless usage of AI tools in ELT.
The survey revealed that more than 78 per cent of students reported frequent use of AI applications like Grammarly, ChatGPT, or Duolingo. These informants reported that their writing fluency, vocabulary learning, and speaking accuracy increased. Qualitative comments analysis showed that students valued AI tools for fast feedback and for increasing their independence in correcting from traditional teacher rule correction. This is consistent with
Students who used the AI tool improved significantly in all four primary skills:
Writing: Users of Grammarly and Write & Improve made less grammatical errors and felt more confident in their academic writing. This is in line with the results of
Speaking & pronunciation: ELSA Speak users experienced 40% improvement in pronunciation on average at the phoneme level. Several students reported that the exposure of AImediated speaking practice decreased their anxiety of speaking in front of a group testifying to the claims of
Most of the interviewed teachers were convinced that the AI tools improved classroom efficiency and facilitated learner autonomy. Ten of 12 teachers indicated that they used Grammarly, ChatGPT, or automated speech technologies to complement their face to face practices. They emphasized AI's potential for individual learning, --whereby "gifted students would be able to work on their own independent tasks, as other students require focused support. But obstacles were training inadequacies, push back by some colleagues and overreliance by students on AIgenerated material. As one teacher observed, "They hand in work that ChatGPT writes for them without being able to comprehend the writing", in reflection of
Personalization: AIbased instruments promoted learner autonomy exposing the learners to realtime and self pacing learning environments. There was also an increased desire among students to learn English outside the classroom.
But institutional readiness varies, and is far from universally in place. Many others didn't have the infrastructure to accommodate fullon AI integration. Some of the major impediments include restricted access to worldclass AI tools, sporadic internet connectivity and no training on digital literacy, according to the teachers. These results are consistent with
The study found an equilibrium between AI utilization and human guidance. Tools such as ChatGPT present huge learning potentials, but also the danger of leading to plagiarism and shallow learning when not properly controlled. Teachers advised us to embed AI literacy in the curriculum and to show students how to interact with the products or services produced by AI in an ethical and critical way. Additionally, content biases in AI tools - especially influenced by Western norms in English - require pedagogical vigilance. As
This study provides some valuable directions of the category regarding AIbased ICT tools in integration with ELT at the universities level of Bangladesh; however, several limitations can be presented. The study was also limited to two public universities in Dhaka. Therefore, the results cannot be generalized to rural communities with substantial infrastructural and pedagogical disparities. Additional institutions would need to be included in future studies to acquire a more accurate understanding of the national landscape. Second, although the mixed method facilitated a comprehensive analysis, the study used selfreported information and may have been affected by information bias. What this study suggested was that students might have exaggerated their involvement or underestimated usage of AI tools for cheating. Third, the tools that researcher investigated in this work were mostly commercial platforms (e.g., Grammarly, ChatGPT, Duolingo), thus are rapidly updating. Therefore, the results may be affected by such versions that were available at the time of data collection and may limit the longterm validity of the study. Fourth, while the researcher took into account teachers' perceptions, the quality of administrators' or policylevel perspectives, which are decisive in institutionalizing AI use curriculum design and teacher training, was not explored. Finally, although AI literacy was indicated as an important consideration, digital competence among the students and faculty was not measured as such. A more focused evaluation on this aspect might guide training and curriculum reform more solidly. Nevertheless, the conclusions could serve as a springboard to further empirical research in AIdriven language learning in lowresource educational settings.
The recommendations for the relevant stakeholders of English language in the context of Bangladesh given based on the findings of this study and its limitations are as follows:
Embed AI Digital Literacy in ELT Courses: Universities need to develop digital and AI literacy components in their English courses. These modules ought to instruct students in how to learn to conscientiously and effectively employ AI tools to enhance language, and not to become inoculated with them. Universities should offer professional development courses to EL teachers to be trained on the pedagogical use of AI like Grammarly, ChatGPT and ELSA Speak. In order to use AI in student centered learning settings, teachers require not only technical competencies but also pedagogical knowledge. Institutionallevel support should be established when the policy facilitates the strategic implementation and deployment of ICT to enhance English Language Teaching (ELT) by ensuring access to subscription tools, upgrading internet services, and promoting effective interdisciplinary collaboration between language departments and technology specialists. Educators should consider inserting assignments that require the use of AI into the curriculum to enhance autonomous learning. For example, students might be instructed to rewrite their essays with AWE technology or participate in conversational practice with AI chatbots on a daily basis. Moreover, the Government and NGOs should collaborate to improve access to AI in rural and nonsponsored institutions. AIpowered language learning can be democratised with mobilebased learning applications and community training workshops.