Using AI to Create ESL Activities for Mixed-Level Classes

Introduction

English as a Second Language (ESL) classrooms are increasingly composed of students with diverse linguistic backgrounds and varying levels of proficiency. From absolute beginners to advanced learners, the complexity of these mixed-level environments poses a significant challenge for educators striving to design lessons that are both inclusive and pedagogically effective. Traditional resources, while useful, often lack the flexibility to accommodate the wide spectrum of abilities present in modern classrooms, leading to uneven engagement, frustration, and reduced learning outcomes.

As educational technologies advance, artificial intelligence (AI) has emerged as a promising solution to these challenges. AI-powered tools offer dynamic, customizable learning experiences that can be tailored to meet the specific needs of individual learners. These technologies allow teachers to automate certain aspects of lesson preparation, personalize content delivery, and assess student performance with greater accuracy and responsiveness. By incorporating AI into the ESL classroom, educators can shift their focus from repetitive administrative tasks to more meaningful instructional interactions.

The integration of AI in language teaching is not merely a matter of convenience; it reflects a shift in how learning is designed and experienced. AI tools such as generative text platforms, intelligent tutoring systems, and adaptive assessment applications can analyze language input, suggest content modifications, and recommend targeted interventions. This capacity to differentiate instruction is particularly valuable in mixed-level classrooms, where one-size-fits-all approaches are typically ineffective.

At the same time, the introduction of AI in ESL education must be approached with thoughtfulness. It requires an understanding of both pedagogical goals and technological limitations. The role of the teacher remains central, not only as a guide and facilitator but also as a critical evaluator of AI-generated content. Teachers must ensure that the materials produced respect the learners’ abilities and support respectful and inclusive classroom practices.

This article aims to provide a comprehensive guide for ESL educators, educational professionals, and interested readers on how AI can be used to create engaging and effective activities in mixed-level ESL classrooms. It will explore the theoretical foundations of mixed-level teaching, offer practical insights into available AI tools, demonstrate strategies for creating adaptive learning experiences, and address ethical considerations surrounding technology use. By balancing academic insight with practical application, this article seeks to empower educators with the knowledge and tools needed to harness AI in ways that enhance, rather than replace, the human dimensions of teaching.

Understanding Mixed-Level ESL Classrooms and the Need for Adaptive Tools

What are mixed-level classes in ESL?

Mixed-level ESL classes are learning environments where students possess a wide range of English language abilities within the same instructional setting. These classes may include learners with different proficiency levels in reading, writing, listening, and speaking, as well as varying educational backgrounds, learning speeds, and experiences with English. Unlike homogeneous classrooms, mixed-level groups challenge teachers to balance competing needs while maintaining lesson coherence and engagement.

Mixed-level classrooms are increasingly common due to factors such as open enrollment policies, adult education programs, and community-based learning initiatives that prioritize access over language grouping. As Lightbown and Spada (2021) observe, these classes are not anomalies but a growing feature of real-world language education that requires strategic instructional planning and flexible teaching practices.

Challenges teachers face when planning lessons for these groups

Planning lessons for mixed-level classes presents several pedagogical challenges. One primary issue is the difficulty of designing activities that are both accessible to lower-level students and stimulating for higher-level learners. Instructors often struggle to provide adequate language input and practice that suits the full range of learners. For example, a listening activity calibrated for intermediate learners may overwhelm beginners and bore advanced students.

Classroom management can also be more complex. Learners working at different paces may require varying levels of attention, which can result in fragmented instruction. Teachers may find themselves unable to monitor progress uniformly or provide timely feedback to all students. According to Gkonou and Miller (2017), the need for individualized support in such settings increases the cognitive and emotional demands on instructors, potentially leading to teacher burnout.

Assessment becomes another hurdle. Grading or evaluating student performance fairly while accounting for differing abilities and progress rates often requires differentiated rubrics and varied forms of testing. Without sufficient tools or support, maintaining motivation and progression across the proficiency spectrum becomes a difficult task.

Importance of differentiation and engagement strategies

Differentiation is a key instructional approach in mixed-level classrooms. It involves modifying content, process, and products of learning to align with students’ language abilities, interests, and learning preferences (Tomlinson, 2014). Effective differentiation may include grouping students strategically, offering scaffolded tasks, or allowing multiple options for demonstrating understanding.

Equally important is the ability to foster respectful engagement across proficiency levels. When learners feel acknowledged and challenged appropriately, their motivation increases. Conversely, if tasks are too easy or too difficult, learners may disengage. Vygotsky’s (1978) concept of the Zone of Proximal Development (ZPD) supports the notion that instruction should target a level just beyond what learners can do independently, providing enough challenge to stimulate learning while offering sufficient support.

However, implementing differentiated instruction in practice is resource-intensive. Teachers need tools that allow them to create, adapt, and deliver activities efficiently and responsively.

Why static resources often fall short

Traditional ESL teaching materials, such as textbooks and printed worksheets, are usually designed for specific proficiency levels. These static resources lack the flexibility required for tailoring lessons in real time. Although publishers may include optional activities or leveled supplements, the content often remains too rigid to accommodate the full spectrum of learners in a mixed-level classroom.

Furthermore, static materials are generally not responsive to learner progress or feedback. They cannot adjust task complexity based on student input, nor do they provide data on student engagement or comprehension. As a result, teachers must often manually adjust tasks or create multiple versions of the same lesson—a time-consuming and sometimes unsustainable approach.

Digital resources have attempted to bridge this gap by offering editable or interactive materials, but even these require considerable teacher input and often lack adaptive intelligence. As Winke and Goertler (2008) point out, many online tools merely digitize traditional resources without leveraging the capabilities of emerging technologies such as AI.

The role of AI-powered solutions

Artificial intelligence offers an opportunity to overcome the limitations of static and semi-dynamic resources. AI systems can process learner data to adjust task difficulty, suggest personalized content, and provide real-time feedback. These tools can generate language activities across skill levels with minimal manual intervention, thereby supporting differentiated instruction at scale.

For example, AI chatbots can engage students in conversation tailored to their proficiency, while text generation tools can create reading passages with adjustable lexical complexity. Automated assessment platforms can provide individualized feedback, allowing students to progress at their own pace without requiring the teacher to manually grade each submission.

By reducing the administrative load and enhancing instructional responsiveness, AI enables teachers to focus more on relational aspects of teaching—mentorship, facilitation, and classroom interaction. This shift aligns with current pedagogical thinking, which emphasizes the teacher’s role in creating meaningful learning experiences rather than merely delivering content.

In conclusion, mixed-level ESL classrooms present complex instructional challenges that require flexible and responsive teaching tools. Static resources, while foundational, are insufficient for addressing the range of learner needs present in such environments. AI-powered technologies hold significant promise for supporting differentiated instruction, improving learner engagement, and enhancing pedagogical efficiency. As the field continues to evolve, educators who are equipped with adaptive tools and informed strategies will be better positioned to meet the needs of all their learners effectively.

Overview of AI Technologies Relevant to ESL Instruction

Key AI technologies used in ESL instruction

Artificial intelligence has become increasingly integrated into educational practice, particularly within English language instruction. Several core AI technologies are pivotal in shaping modern ESL classrooms. These include natural language processing (NLP), machine learning, chatbots, and generative AI.

Natural language processing refers to the ability of computers to interpret and generate human language. This technology powers applications that analyze grammar, check pronunciation, and provide feedback on written and spoken English. Tools like Grammarly and Microsoft Editor rely heavily on NLP to evaluate syntax, semantics, and writing style in real time (Jurafsky & Martin, 2023).

Machine learning is another foundational technology, enabling software systems to improve performance through experience. In the ESL context, machine learning allows platforms to adapt to student behavior, personalize content, and predict areas where learners might struggle. This adaptive functionality supports more efficient language acquisition by delivering appropriate challenges based on ongoing performance (Zawacki-Richter et al., 2019).

Chatbots, driven by rule-based algorithms or more sophisticated machine learning models, simulate human conversation. They are used in ESL instruction to provide real-time practice in speaking and writing. These bots can role-play different conversational scenarios, helping learners improve fluency and confidence in a low-pressure environment (Fryer et al., 2020).

Generative AI, including models such as ChatGPT, can produce original text, answer questions, summarize content, and even simulate dialogue. For ESL teachers, this technology offers a powerful means of generating differentiated reading passages, vocabulary activities, writing prompts, and grammar exercises with minimal manual input.

How these technologies interact with language learning platforms

AI tools do not operate in isolation; they are typically embedded within larger digital ecosystems, such as language learning platforms or classroom management systems. These platforms integrate AI technologies to enhance content delivery, automate feedback, and support communication between teachers and students.

For example, Duolingo employs machine learning to adapt question difficulty and suggest review exercises based on individual learner performance. The platform uses data from millions of learners to refine its algorithms continually, creating a responsive learning environment that adjusts to user needs (Settles & Meeder, 2016).

Similarly, Cambridge One and Oxford English Hub have incorporated AI-driven diagnostic assessments that inform instructional planning. These assessments evaluate learner progress and generate reports that help teachers identify skill gaps and tailor instruction accordingly. These platforms combine user interaction data with AI to present dashboards, analytics, and content suggestions that support decision-making at both the classroom and institutional levels.

AI integration also supports multilingual learners by offering translation services, speech recognition tools, and voice-assisted applications, such as Google’s Read Along and Microsoft’s Immersive Reader. These functionalities reduce barriers to access and enable learners to work independently while reinforcing the teacher’s role as a guide and facilitator.

Free and paid AI tools available for teachers

Teachers today have access to a wide range of AI-powered tools, many of which are designed to support lesson planning, material generation, and formative assessment. These tools can be categorized as either free or subscription-based, each with distinct capabilities and limitations.

Free tools:

• ChatGPT (Free version by OpenAI): Capable of generating writing prompts, grammar drills, comprehension questions, and role-play dialogues.
• QuillBot: Offers paraphrasing assistance and summarization, useful for vocabulary building and text simplification.
• Rewordify: Simplifies difficult English text, making reading passages accessible to learners at various levels.
• Google Translate with NLP enhancements: Provides translations with contextual understanding, aiding in vocabulary acquisition and comprehension.
• Quizlet AI: Uses AI to generate flashcards and practice quizzes from uploaded content, facilitating personalized review sessions.

Paid tools:

• Grammarly Premium: Offers detailed grammar feedback, tone suggestions, and writing enhancements suitable for intermediate to advanced learners.
• Wordtune: Provides rewriting suggestions based on user intent (formal, casual, expanded), aiding in the development of academic and conversational writing.
• Texthelp Read&Write: Combines text-to-speech, translation, and vocabulary support to aid learners with reading and comprehension.
• Elsa Speak: Focused on pronunciation improvement through AI-driven feedback, particularly useful for speaking practice.
• Khanmigo (AI by Khan Academy): An experimental AI teaching assistant that can help with language explanations and grammar tutorials.

Each tool varies in terms of accessibility, ease of use, and pedagogical alignment. Free tools are often limited in customization and depth, while paid tools typically offer more robust features, such as analytics, content management, and adaptive learning.

Criteria for selecting AI tools for pedagogical purposes

Selecting the right AI tools requires a careful consideration of both pedagogical goals and classroom contexts. The following criteria can assist educators in evaluating AI applications for ESL instruction:

1. Alignment with learning outcomes: Tools should support specific instructional goals, whether vocabulary acquisition, grammar reinforcement, speaking fluency, or writing development.
2. Adaptability across proficiency levels: Given the variability in mixed-level classrooms, tools should offer flexible settings to accommodate different learner needs.
3. User interface and accessibility: Teachers and students should be able to navigate the tool easily, with minimal training. Mobile compatibility and offline access are added advantages.
4. Data privacy and security: Tools should comply with institutional and legal standards for protecting student information, particularly in school or public-sector contexts (Slade & Prinsloo, 2013).
5. Integration capabilities: Tools that can be embedded within Learning Management Systems (LMS), such as Google Classroom or Moodle, provide smoother implementation and consistency across platforms.
6. Feedback mechanisms: Effective tools should offer immediate, meaningful feedback that guides learners toward improved performance, whether through grammar correction, speech evaluation, or comprehension prompts.

Ultimately, AI technologies must be viewed as instructional aids rather than replacements for thoughtful teaching practice. Their effectiveness lies in how well they support differentiated instruction, reduce teacher workload, and provide engaging, level-appropriate experiences for students.

Designing AI-Enhanced Activities for Mixed-Level Learners

AI-based content generation for different proficiency levels

One of the key advantages of artificial intelligence in ESL instruction is its ability to generate content tailored to various language proficiency levels. AI tools can produce materials that match a learner’s current skills while supporting progression through scaffolded learning. For example, large language models such as ChatGPT or Claude.ai can be prompted to generate texts at specific CEFR levels (A1 to C2), making them highly adaptable to classroom needs.

With careful input, teachers can use AI to produce reading texts, cloze activities, or writing prompts appropriate for beginner, intermediate, or advanced students. For instance, the same prompt, “Describe your favorite day of the week,” can yield three distinct outputs depending on the complexity requested. A beginner-level output may use simple present tense and common vocabulary, while an advanced version could incorporate figurative language and varied sentence structures.

Additionally, tools like TextEngine.io or Rewordify can simplify authentic texts, such as news articles or short stories, based on target vocabulary bands. This capability allows all learners to engage with the same topic but at levels they can comprehend and respond to effectively, supporting balanced classroom discussions and collaborative work.

Scaffolded task creation: vocabulary, grammar, and communicative skills

Effective mixed-level instruction relies on scaffolding—an approach that gradually builds learners’ skills through structured support. AI tools can be used to design activities that follow this model, from foundational drills to open-ended tasks. Teachers can prompt AI platforms to produce sequences of tasks that develop vocabulary, grammar, and communication skills incrementally.

For vocabulary instruction, platforms like Quizlet AI or Wordwall can create tiered flashcards, matching exercises, and games that adjust based on the learner’s proficiency. These tools often allow the teacher to set difficulty levels or input target word lists, which the system then uses to generate customized learning paths.

Grammar instruction also benefits from AI-powered scaffolding. For example, Grammarly and Write & Improve (by Cambridge) offer real-time feedback on sentence construction, verb forms, and article usage. Such feedback supports immediate learning while allowing students to revise and improve iteratively. Instructors can also use ChatGPT to generate controlled grammar practice activities, such as fill-in-the-blank sentences or transformation exercises, adjusted for various grammar points and complexity levels.

Communicative competence can be fostered through AI chatbots or dialogue generators. For example, teachers can use AI to simulate real-life scenarios (e.g., ordering food, giving directions) by creating scripts with blanks that students fill based on role-play contexts. These scripts can be designed at multiple levels of complexity, allowing beginners to focus on key phrases while advanced learners expand with elaboration or polite forms.

Examples of differentiated listening, reading, speaking, and writing activities

AI tools enable the creation of differentiated activities across all four language skills—listening, reading, speaking, and writing—ensuring that learners engage meaningfully regardless of their proficiency level.

Listening:
Teachers can use AI to generate or select audio materials suited to different comprehension levels. Applications like Speechling or FluentU offer AI-curated listening exercises with adjustable playback speeds and transcripts. For example, a topic like “travel” can be explored using simple audio for beginners (e.g., “I go to the beach”), while advanced learners tackle complex dialogues or podcasts with multiple speakers and accents.

AI can also be used to create listening comprehension questions at varying depths. Teachers may use tools such as ChatGPT to generate multiple-choice or open-ended questions for a single audio clip, varying the question types and vocabulary complexity to match different learner levels.

Reading:
Using AI-generated texts, educators can prepare a reading activity on a shared topic, such as environmental issues, and produce three different versions: one with basic sentence structure and common vocabulary, one with moderate complexity and transition signals, and one featuring academic vocabulary and varied syntax. Paired with AI-generated comprehension questions or summarizing tasks, this approach allows mixed-level learners to work independently or in level-based groups.

Speaking:
AI-based conversation simulators, such as TalkPal or AI English Coach, enable learners to practice speaking through guided prompts or free-form interaction. These tools often provide real-time speech recognition and pronunciation feedback, which is particularly helpful for students who may be hesitant to speak in front of peers. Teachers can assign speaking tasks based on proficiency levels—beginners might respond to picture prompts, while advanced learners engage in AI-mediated debates or problem-solving activities.

Writing:
Writing practice can be greatly enhanced using AI-generated prompts that target specific structures and themes. For example, students can be asked to describe a place, express an opinion, or write a letter. Grammarly or Write & Improve can then provide structured feedback on grammar, style, and coherence. Teachers can further support mixed-level learning by assigning tiered writing prompts, such as describing a family member (basic) versus narrating a personal growth experience (advanced).

AI can also assist in the revision process. By comparing multiple drafts with AI feedback, learners improve their metacognitive awareness of grammar, coherence, and tone—an important step in developing independent writing skills (Ranalli, Yamashita, & Henstock, 2021).

Encouraging learner autonomy using AI tools

In mixed-level classrooms, fostering learner autonomy is essential. Students benefit when they can self-select resources and manage their own progress. AI tools contribute to this goal by offering feedback, pacing, and scaffolding without teacher intervention.

Platforms such as LingQ or BBC Learning English’s chatbot support self-paced learning, where students choose topics and levels that suit their interests and abilities. These tools encourage exploration and repeated practice, which are critical for language retention.

Moreover, learners can use AI as a private tutor. For example, they may use ChatGPT to request sentence corrections, vocabulary explanations, or grammar rules. When guided appropriately, this interaction promotes reflective learning and builds learner confidence in experimenting with language. However, educators should provide training on how to use these tools critically and responsibly, emphasizing that AI is a supplement, not a substitute, for thoughtful instruction and interaction.

Structured AI-integrated learning journals are another method for promoting autonomy. Students can reflect on their learning, ask AI for paraphrasing help, or brainstorm ideas before writing. Over time, this practice encourages the development of independent language learning strategies.

Instructors can also use AI-generated progress tracking tools, such as those embedded in language apps or LMS platforms, to encourage learners to monitor their own development. These dashboards provide visual feedback on performance and areas for improvement, fostering a growth mindset and accountability.

Managing Student Engagement and Motivation with AI

Gamification and personalization through AI

Student engagement in ESL classrooms, particularly those with mixed levels of proficiency, can be difficult to sustain without deliberate and dynamic instructional strategies. Gamification, powered by artificial intelligence, offers a pathway to increase learner motivation by transforming conventional tasks into interactive experiences. AI enhances gamification by analyzing learner performance and adapting activities to individual progress, thus providing a sense of challenge and achievement that sustains attention and effort.

AI-supported platforms such as Duolingo and Classcraft employ reward systems, points, levels, and instant feedback to encourage sustained participation. These features are not arbitrary; they are driven by adaptive algorithms that track user behavior and recommend tasks calibrated to keep learners in what Csikszentmihalyi (1990) described as the “flow zone,” where tasks are neither too easy nor too difficult. This state of engagement fosters deeper learning and greater persistence.

Personalization plays a parallel role in motivation. AI tools like LingQ or Sora by OverDrive allow learners to select texts and topics of personal interest while maintaining appropriate language levels. The alignment of content with learners’ interests and needs enhances their intrinsic motivation, making them more likely to invest time and effort in completing tasks (Dörnyei & Ushioda, 2021).

Moreover, AI’s ability to remember learner preferences and prior responses allows platforms to create learning pathways that feel personally tailored. This customization can be especially motivating for learners who feel left behind or under-challenged in mixed-level groupings, ensuring each student encounters content that is engaging and meaningful.

Real-time feedback and AI-powered formative assessment

Feedback is a cornerstone of effective language learning. However, in mixed-level classrooms, the challenge lies in delivering timely and appropriate feedback to all learners. AI tools address this gap by offering real-time feedback that is both immediate and level-sensitive. This responsiveness not only reduces the teacher’s workload but also empowers students to take corrective actions while the task is still relevant.

For writing tasks, platforms such as Grammarly, Write & Improve, and Quill provide automated feedback on grammar, vocabulary usage, coherence, and even tone. Importantly, these tools offer explanations and suggestions rather than just corrections, facilitating a deeper understanding of language use. According to Bitchener and Ferris (2012), such formative feedback supports the development of linguistic accuracy and learner self-regulation when integrated into ongoing instruction.

In speaking practice, applications like Elsa Speak use AI-driven speech recognition to analyze pronunciation, intonation, and rhythm, offering learners insights into their oral proficiency. These apps enable learners to repeat and refine their speech in real-time, promoting confidence and fluency.

Teachers can also use AI-enabled learning management systems that track learner progress, analyze patterns of errors, and present individualized performance reports. These systems allow for differentiated formative assessment, providing students with clear, actionable feedback while informing teachers about instructional needs.

AI-generated content for creative language tasks

While engagement can be increased through structure and feedback, creativity also plays a crucial role in motivating learners. AI tools can help foster creativity by generating prompts, scenarios, and dialogue scripts that encourage learners to experiment with language in imaginative ways.

For instance, AI text generators like ChatGPT can create story starters, character profiles, or debate topics based on user-defined parameters. Teachers might prompt the AI to generate a mysterious first paragraph of a short story and then ask students to continue the narrative in groups or individually. Tasks like these promote expressive language use and push learners beyond formulaic structures.

Roleplay activities can also benefit from AI support. Tools like Scenario by Immerse or AI Dungeon allow learners to explore fictional or real-life scenarios, such as job interviews or travel situations, in a guided but open-ended format. These activities promote communicative competence and are easily modified to suit beginner through advanced learners.

Importantly, creative tasks generated by AI can be revisited and revised, providing additional opportunities for formative feedback and self-assessment. Teachers can scaffold the activity using rubrics or learner checklists, ensuring that creativity is supported by clear language learning goals.

Encouraging respectful peer interaction with AI support

Peer interaction remains essential in language development, especially in communicative tasks that develop fluency and social use of language. However, managing peer interactions in mixed-level classrooms can be challenging. AI can support respectful and productive peer engagement by helping learners prepare, rehearse, and reflect on their interactions.

For example, before a group discussion, learners can use AI tools to draft ideas, generate vocabulary lists, or practice pronunciation. Chatbots and conversation simulators allow students to rehearse interactions privately, boosting their readiness and confidence. After the activity, they can review transcripts or summaries created by AI tools, encouraging self-reflection and respectful peer evaluation.

Collaborative platforms such as Google Docs, when combined with AI assistants like Grammarly or Wordtune, allow learners to co-write and revise texts with AI-supported suggestions. These shared tasks promote cooperation and mutual support. To avoid imbalance in participation, instructors can assign roles (e.g., editor, researcher, writer) and use AI-generated checklists to guide group work.

In multilingual and mixed-level settings, AI translation tools like DeepL or Microsoft Translator can serve as a temporary support mechanism, helping learners understand instructions or communicate more clearly. However, teachers should guide learners in using these tools responsibly, emphasizing the importance of building independent communicative competence over time (Bozkurt et al., 2021).

Respectful interaction is further promoted when AI-generated content models courteous, constructive language. Teachers can prompt AI tools to produce dialogue samples or feedback phrases that students can emulate. For example, learners may be shown how to express disagreement politely or offer feedback constructively. This not only aids language learning but also fosters a respectful classroom environment.

AI for Assessment in Mixed-Level ESL Classrooms

Formative vs summative assessment with AI tools

Assessment plays a crucial role in ESL instruction, particularly in mixed-level classrooms where understanding individual learner needs is essential. AI technologies have introduced new possibilities for conducting both formative and summative assessments efficiently and with greater responsiveness to learner variation.

Formative assessment refers to ongoing evaluations that inform instruction and guide learning during the instructional process. AI tools facilitate formative assessment by providing real-time feedback, monitoring student progress, and recommending instructional adjustments. For example, platforms such as Write & Improve and Grammarly offer instant feedback on grammar, coherence, and vocabulary usage, allowing learners to revise and resubmit their work. This continuous feedback loop supports learner development and enables teachers to identify patterns in student errors (Black & Wiliam, 2009).

In contrast, summative assessment typically occurs at the end of an instructional unit or course and is designed to evaluate overall achievement. AI-powered systems can be used to automate grading and generate consistent results. Tools like ExamSoft and Moodle’s AI-based quiz modules allow for automatic marking of objective questions, reducing teacher workload while maintaining reliability. AI also contributes to summative writing assessments through rubric-based scoring and linguistic feature analysis, although such systems should be reviewed critically to ensure alignment with instructional goals and fairness in evaluation (Poe et al., 2021).

Adaptive testing and proficiency tracking

One of the most impactful applications of AI in assessment is adaptive testing. Adaptive tests adjust the difficulty of questions in real time based on the learner’s responses. This approach offers a more precise measure of proficiency by focusing on the learner’s actual level rather than presenting a uniform set of questions to all students.

The Duolingo English Test (DET) is a notable example of an AI-driven adaptive test. It evaluates reading, writing, listening, and speaking through tasks that adapt dynamically. As a student answers correctly, the system presents more challenging questions; incorrect answers trigger easier ones. This mechanism ensures that learners are neither overwhelmed nor under-challenged, providing a fair assessment of ability across a wide range of proficiency levels (Yan & Zhang, 2021).

Proficiency tracking through AI involves the continuous collection and analysis of learner performance data. Tools embedded within learning management systems, such as Khan Academy or Google Classroom’s AI analytics, track progress over time and highlight skill gaps. These insights help teachers make informed decisions about grouping, targeted instruction, and resource allocation. Teachers can access dashboards that visualize trends and flag learners who may need additional support or challenge.

Creating custom quizzes with AI

Creating assessments tailored to different learner levels can be time-consuming. AI streamlines this process by enabling teachers to generate custom quizzes that reflect learners’ abilities and instructional objectives.

AI-powered platforms like ChatGPT, Quizizz AI, and Conker by Microsoft allow educators to input text, themes, or vocabulary lists and automatically receive multiple-choice questions, cloze tests, true/false items, or short-answer questions. Teachers can also specify target CEFR levels or Bloom’s taxonomy objectives to ensure alignment with their curriculum. For instance, a teacher can prompt ChatGPT to create comprehension questions for a B1-level reading passage, followed by inference or vocabulary exercises for more advanced learners.

These tools also permit the quick generation of parallel versions of the same quiz for different proficiency bands. For example, a grammar quiz on present perfect tense might feature more contextualized sentence completions for advanced learners and controlled fill-in-the-blanks for beginners. This allows all learners to engage with the same grammatical structure at a pedagogically appropriate level.

Some platforms offer integration with classroom software, enabling teachers to administer AI-generated quizzes directly through LMS tools like Moodle or Canvas. The ability to generate, edit, and deliver assessments in a digital environment enhances instructional flexibility and responsiveness.

Feedback strategies for multiple proficiency bands

Effective feedback is essential to assessment, especially in classrooms with mixed proficiency levels. AI supports this process by generating differentiated feedback that caters to learners’ unique strengths and areas for improvement.

For written work, AI tools such as Write & Improve or Wordtune provide comments that are sensitive to the learner’s level. Beginners may receive feedback on article usage and sentence structure, while advanced students are prompted to improve cohesion, precision, and tone. These distinctions ensure that all learners receive feedback that is both accessible and challenging.

For spoken responses, applications like Elsa Speak and Microsoft’s Reading Progress provide detailed phonetic analysis and pronunciation feedback. Learners receive visual indicators showing which sounds were mispronounced and suggestions for improvement. These systems often allow teachers to adjust feedback sensitivity to accommodate different proficiency levels.

Teachers can enhance AI-generated feedback by embedding rubrics that define performance standards for each level. Some platforms offer customizable rubrics that can be adapted to classroom goals. Providing learners with level-specific success criteria promotes transparency and encourages reflection. According to Hattie and Timperley (2007), effective feedback should answer three questions: “Where am I going?” “How am I going?” and “Where to next?” AI tools can help address all three when used in conjunction with clear rubrics and learning targets.

AI also supports peer feedback in structured ways. For example, platforms like Peergrade and Eduflow allow students to review each other’s work with guidance from AI-generated prompts or rating scales. These systems ensure that peer reviews remain respectful, constructive, and level-appropriate, even in mixed-ability groups.

By incorporating AI into the feedback process, teachers are better equipped to support learners individually, reduce grading time, and focus on instructional design and facilitation. The key lies in selecting tools that align with classroom goals and maintaining a human-centered approach to interpretation and guidance.

Addressing Ethical and Practical Considerations in Using AI

Data privacy and student safety

As artificial intelligence becomes increasingly integrated into ESL instruction, protecting student data and ensuring digital safety are paramount. AI-powered tools often collect and process vast amounts of learner information, including login credentials, performance data, audio recordings, written submissions, and behavioral patterns. This data is essential for powering adaptive learning features but presents significant ethical challenges when not properly managed.

Many educational technology platforms operate under cloud-based systems, raising concerns about data storage, access, and sharing. According to the International Society for Technology in Education (ISTE, 2022), educators and institutions must ensure that the AI tools they adopt comply with international data protection regulations such as the General Data Protection Regulation (GDPR) and the Family Educational Rights and Privacy Act (FERPA) in the United States.

Teachers and administrators should investigate whether a platform stores data locally or in third-party servers and confirm whether it allows user consent before data collection begins. Transparent privacy policies, data encryption, and user control over data deletion are essential features of ethical AI tools. Without these safeguards, students, particularly minors, may be exposed to risks including data misuse, unauthorized surveillance, or targeted advertising (Williamson & Hogan, 2020).

In practice, schools must also consider how they manage student accounts, whether through institutional email addresses, single sign-on systems, or third-party logins. Training teachers to understand basic cybersecurity principles, such as password security and safe browsing habits, also plays a vital role in promoting student safety.

Bias in AI-generated materials

Another pressing concern in AI use is the potential for algorithmic bias. AI systems are trained on large datasets that may contain imbalances, stereotypes, or inaccuracies. When these datasets reflect historical or cultural biases, the resulting outputs may reproduce or reinforce them, even in educational contexts.

In ESL instruction, this may manifest in AI-generated texts that overrepresent certain cultural norms, marginalize others, or offer narrow perspectives on social issues. Studies have shown that generative AI models can produce outputs that favor dominant worldviews or linguistic norms, particularly those aligned with North American or British English conventions (Bender et al., 2021).

For example, an AI-generated roleplay might suggest a scenario such as “ordering a steak at a restaurant” without considering regional dietary habits, or it may overuse idiomatic expressions that are unfamiliar or inappropriate for learners in specific contexts. While such issues may seem minor, they can limit learner engagement or present a skewed view of English usage.

To mitigate these concerns, teachers should take an active role in reviewing and adapting AI-generated content. This includes checking for cultural relevance, age appropriateness, and pedagogical alignment. Teachers may also prompt AI tools with specific constraints or examples to produce more balanced outputs. Educators using AI models like ChatGPT or Claude can refine prompts to include local examples or specify tone and content guidelines to ensure materials remain respectful and pedagogically sound.

Teacher responsibility and tool transparency

AI can serve as a powerful instructional aid, but it is not a substitute for informed and reflective teaching. Teachers must remain the final decision-makers when integrating AI-generated content into their classrooms. Understanding the strengths and limitations of AI tools is crucial to ensuring they support rather than undermine instructional integrity.

Transparency in AI functionality is also a key concern. Many AI-powered platforms operate as “black boxes,” offering little information about how decisions are made or how algorithms process learner data. For example, a grammar correction tool might suggest a change without explaining the rationale, which can hinder rather than promote language learning.

Educators should prioritize tools that offer explainability—that is, the ability to clarify how and why certain suggestions or adaptations are made. Tools that offer user-friendly explanations, citations, or error codes enhance trust and support learner autonomy. Teachers should also communicate openly with students about the role of AI in their learning and encourage them to question and reflect on the outputs they receive.

As Luckin et al. (2016) emphasize, the teacher’s role must evolve to include AI literacy—not just how to use the tools, but how to evaluate, interpret, and manage them within pedagogical frameworks. Professional development initiatives that include AI awareness and ethical use guidelines are therefore essential.

Sustainability of AI integration in resource-limited environments

While AI tools offer substantial pedagogical benefits, their effective integration assumes access to stable internet connections, reliable devices, and digital literacy among users. In many settings, these conditions are not guaranteed, particularly in public education systems or rural areas.

Schools operating under budget constraints may find it difficult to adopt premium AI services, provide adequate hardware, or support ongoing technical training. Additionally, bandwidth-heavy applications like voice recognition software or video-based AI tutors may not function well in low-resource environments.

To address these limitations, educators can consider lightweight, offline-compatible, or open-source AI tools that do not require continuous internet access. For instance, downloading AI-generated materials for offline use or relying on browser-based versions of tools rather than full software installations can reduce the technological burden.

Moreover, institutions should assess the long-term viability of AI integration by considering licensing costs, maintenance requirements, and teacher support structures. Partnerships with nonprofit educational organizations or participation in government-supported edtech initiatives may help reduce costs and improve access.

Ultimately, sustainable integration depends not only on infrastructure but on pedagogical adaptability. Teachers can blend traditional and AI-enhanced methods to ensure that no learner is disadvantaged due to technical limitations. Hybrid lesson designs, where AI tools supplement rather than dominate instruction, can offer a balanced and more accessible approach.

Future Directions: AI and the Evolving Role of the ESL Teacher

How AI is reshaping teacher roles

The introduction of artificial intelligence into ESL instruction is not merely changing the tools teachers use; it is reshaping the very nature of their professional roles. With AI systems capable of automating grammar correction, generating practice materials, and offering real-time feedback, the traditional image of the teacher as the sole source of content and evaluation is shifting. Instead, teachers are increasingly positioned as learning designers, decision-makers, and facilitators of complex language learning environments.

AI can take on many of the routine, repetitive tasks that once consumed large portions of instructional time, such as correcting low-stakes writing, generating vocabulary lists, or creating quizzes. This shift allows teachers to focus more fully on higher-order instructional strategies, such as analyzing learner needs, moderating discussions, and providing personalized coaching. According to Holmes et al. (2019), AI’s growing influence in education will lead to more collaborative roles between humans and machines, where teachers guide and oversee the meaningful application of intelligent systems.

Balancing automation with human pedagogy

While AI can enhance efficiency and personalization, it cannot replicate the full range of human pedagogical skills. Emotional intelligence, cultural awareness, ethical judgment, and the ability to form trusting relationships with students remain central to effective teaching. The danger of over-relying on AI lies in diminishing these critical human dimensions, particularly in language learning, which depends heavily on social interaction, empathy, and contextual understanding.

In the ESL classroom, the teacher’s role involves more than delivering content—it includes managing classroom dynamics, interpreting student emotions, and adapting instruction based on nuanced observations. As Selwyn (2019) emphasizes, AI can supplement but not replace the teacher’s capacity to exercise professional discretion and emotional responsiveness. For example, while a chatbot may simulate conversation, it lacks the awareness to recognize when a student is frustrated or disengaged and respond accordingly.

Balancing AI automation with human pedagogy requires thoughtful integration. Teachers must determine when and how to deploy AI tools so that they enhance learning rather than displace meaningful teacher-student interaction. This involves making informed choices about which technologies to use, for what purposes, and in which instructional contexts.

The teacher as facilitator, curator, and mentor in AI-supported classrooms

In AI-enhanced ESL environments, the teacher takes on a more dynamic and multifaceted role. As facilitators, teachers guide students through learning experiences that involve both AI-driven and traditional resources. They help students interpret AI-generated feedback, set learning goals, and monitor their own progress. This role is essential for ensuring that students understand how to use AI tools productively and critically.

As curators, teachers select and adapt AI-generated materials to align with curricular goals and student needs. Rather than relying on AI outputs at face value, they review, refine, and sequence materials to ensure pedagogical coherence. This curation process includes filtering out content that is culturally irrelevant or too complex, and modifying tasks to suit particular learner groups.

Teachers also act as mentors, providing the human support that learners need to build confidence and develop motivation. They help students navigate challenges, celebrate successes, and foster a respectful learning environment. This mentorship is particularly valuable in mixed-level classrooms, where learners may progress at different rates and require encouragement tailored to their individual paths.

In these roles, teachers remain central to the educational process, not in spite of AI, but because of their capacity to humanize it. As Luckin (2018) suggests, technology must be guided by informed professionals who understand the learning process and prioritize the well-being of their students.

Preparing for continuous professional development in AI literacy

To fulfill these evolving roles, ESL educators must engage in continuous professional development focused on AI literacy. This includes not only technical skills, such as learning to prompt generative models effectively or interpret learning analytics, but also critical thinking about the ethical and pedagogical implications of AI in language education.

Professional development programs should address questions such as: How can AI be integrated meaningfully into existing curricula? What are the risks of AI overuse? How can teachers support students in becoming responsible users of AI tools?

Organizations like TESOL International Association and the British Council have begun offering webinars, courses, and publications on the use of AI in language education. Teachers can also benefit from peer learning, online forums, and collaborative action research that explores best practices for AI integration.

Moreover, institutions and policymakers must support teachers in this transition by providing access to training, resources, and time for experimentation. Without institutional support, even the most innovative tools may fail to produce meaningful outcomes in the classroom.

Conclusion

The integration of artificial intelligence into English as a Second Language instruction marks a significant advancement in how educators approach the challenges of mixed-level classrooms. Throughout this article, we have explored the multifaceted ways in which AI can enhance lesson planning, support differentiated instruction, increase learner engagement, and streamline assessment. These tools offer concrete solutions to longstanding problems in language education, particularly the difficulty of addressing varied learner needs within a single instructional setting.

AI technologies such as natural language processing, generative text tools, and adaptive learning systems enable teachers to produce content tailored to a wide range of proficiency levels with greater efficiency and precision. Whether generating scaffolded reading tasks, personalized writing prompts, or adaptive grammar exercises, AI serves as a responsive assistant that reduces the time and effort required to prepare and deliver effective instruction.

Moreover, AI supports student motivation through gamification and personalized learning paths, providing real-time feedback that is essential for learner progress. The capacity of AI to offer immediate and level-appropriate feedback allows students to engage more deeply with language tasks and develop greater confidence in their abilities. These features are especially valuable in mixed-level classrooms, where learners often experience uneven access to support and practice opportunities.

Assessment, too, benefits significantly from AI integration. Automated and adaptive testing tools enable ongoing proficiency tracking and help teachers make informed decisions based on real-time data. AI-generated quizzes and diagnostic activities allow for more nuanced evaluation, promoting accurate understanding of student needs without overwhelming instructional time.

However, the successful use of AI in ESL education depends not only on the capabilities of the technology itself but on how thoughtfully it is integrated. As emphasized throughout this article, teachers remain central to the instructional process. AI cannot replace the human capacity for mentorship, cultural sensitivity, or pedagogical judgment. Instead, it should be understood as a tool that augments the teacher’s role, making it possible to focus more on meaningful interaction and learner development.

Mindful adoption of AI also requires attention to ethical considerations such as data privacy, algorithmic bias, and digital access. Teachers and institutions must approach AI with critical awareness, ensuring that its use promotes fair, respectful, and pedagogically sound learning environments.

Looking ahead, the role of the ESL teacher is set to evolve. With appropriate training and professional development in AI literacy, educators can become facilitators, curators, and mentors in technology-enhanced classrooms. Collaboration among teachers, researchers, and technology developers will be essential to ensure that AI tools are aligned with the real needs of learners.

Ultimately, the goal is not to automate education but to enrich it. When used wisely, AI offers a pathway to more engaging, effective, and responsive ESL instruction—one that respects the complexity of language learning and the unique journeys of every student.

References

Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 610–623. https://doi.org/10.1145/3442188.3445922

Bitchener, J., & Ferris, D. R. (2012). Written corrective feedback in second language acquisition and writing. Routledge. https://doi.org/10.4324/9780203832400

Black, P., & Wiliam, D. (2009). Developing the theory of formative assessment. Educational Assessment, Evaluation and Accountability, 21(1), 5–31. https://doi.org/10.1007/s11092-008-9068-5

Bozkurt, A., Karadeniz, A., & Koçdar, S. (2021). Using AI-powered tools in language education: Opportunities and challenges. Asian Journal of Distance Education, 16(2), 66–77. http://www.asianjde.org/ojs/index.php/AsianJDE/article/view/603

Cambridge English. (n.d.). Write & Improve. Cambridge University Press. https://writeandimprove.com

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. Harper & Row.

Dörnyei, Z., & Ushioda, E. (2021). Motivation and second language acquisition (2nd ed.). Routledge. https://doi.org/10.4324/9780367857056

Duolingo English Test. (2023). Duolingo English Test. https://englishtest.duolingo.com

FluentU. (2023). Language learning through real-world videos. https://www.fluentu.com

Fryer, L. K., Ainley, M., Thompson, A., Gibson, A., & Sherlock, Z. (2020). Stimulating and sustaining interest in a language course using chatbots. Computers & Education, 149, 103818. https://doi.org/10.1016/j.compedu.2019.103818

Gkonou, C., & Miller, E. R. (2017). Caring and emotional labour: Language teachers’ engagement with anxious learners in private language school classrooms. Language Teaching Research, 21(4), 446–465. https://doi.org/10.1177/1362168816686984

Heffernan, N., & Otoshi, J. (2015). Computer-based testing and feedback: Perspectives on the potential for automated assessment and tracking in language classrooms. Language Learning & Technology, 19(1), 116–136. https://www.lltjournal.org/item/2908

Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education: Promises and implications for teaching and learning. Center for Curriculum Redesign. http://curriculumredesign.org/wp-content/uploads/AI-in-Education-Promises-and-Implications.pdf

International Society for Technology in Education. (2022). Ethical considerations for AI in education: A guide for educators and leaders. https://www.iste.org/areas-of-focus/artificial-intelligence

Jurafsky, D., & Martin, J. H. (2023). Speech and language processing (3rd ed.). https://web.stanford.edu/~jurafsky/slp3/

Lightbown, P. M., & Spada, N. (2021). How languages are learned (5th ed.). Oxford University Press.

Luckin, R. (2018). Machine learning and human intelligence: The future of education for the 21st century. UCL Institute of Education Press.

Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed: An argument for AI in education. Pearson Education. https://www.pearson.com/en-us/about/news-events/news/2016/01/intelligence-unleashed–an-argument-for-ai-in-education.html

Quizlet. (2023). Introducing Magic Notes and AI features. https://quizlet.com/blog/introducing-magic-notes-and-ai-features

Ranalli, J., Yamashita, T., & Henstock, B. (2021). Does Grammarly help students to improve their writing? An investigation of the effects of an AI-based writing assistant. Journal of Second Language Writing, 52, 100816. https://doi.org/10.1016/j.jslw.2021.100816

Selwyn, N. (2019). Should robots replace teachers? AI and the future of education. Polity Press.

Settles, B., & Meeder, B. (2016). A trainable spaced repetition model for language learning. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 1848–1858. https://aclanthology.org/P16-1175/

Shintani, N., & Ellis, R. (2013). The comparative effect of direct written corrective feedback and metalinguistic explanation on learners’ explicit and implicit knowledge of the English indefinite article. Journal of Second Language Writing, 22(3), 286–306. https://doi.org/10.1016/j.jslw.2013.03.011

TESOL International Association. (2023). TESOL AI and education resources. https://www.tesol.org

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press.

Williamson, B., & Hogan, A. (2020). Commercialisation and privatisation in/of education in the context of Covid-19. Education International Research. https://ei-ie.org/en/detail/16710/commercialisation-and-privatisation-in/of-education-in-the-context-of-covid-19

Winke, P., & Goertler, S. (2008). Did we forget someone? Students’ computer access and literacy for CALL. CALICO Journal, 25(3), 482–509. https://doi.org/10.1558/cj.v25i3.482-509

Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education, 16(1), 39. https://doi.org/10.1186/s41239-019-0171-0