AI to the rescue: Exploring the potential of ChatGPT as a teacher ally for workload relief and burnout prevention

This research investigates the effectiveness of AI generative ChatGPT as a teacher assistant to reduce workload and prevent burnout in secondary schools. Lesson planning and content development were significant contributors to teacher burnout. In response, ChatGPT was tested with tailored queries for English, science, and math subjects, utilising an explanatory research approach to assess ChatGPT’s capabilities in personalised planning and content development, given that there is limited available information around this topic. The study highlights ChatGPT’s benefits in personalised planning through task-specific prompts and AI-human collaboration. Aligned with UAE’s AI integration objectives, the study emphasises balanced use and educational reform potential. Integrating AI tools optimises teacher planning, enhances instructional support, and refines resource allocation, contributing to AI’s academic potential while stressing burnout mitigation’s importance for educational advancement.

Consequently, exploring teacher burnout and its associated factors has attracted considerable attention from educational researchers and policymakers (Saloviita & Pakarinen, 2021).Gaining insight into the factors contributing to teacher burnout can inform decisions and actions to enhance school environments, promote teacher job satisfaction, and deliver high-quality education (Granziera et al., 2021).
Teaching has undergone significant transformations in the 21st century, primarily driven by technological advancements like artificial intelligence (AI) (Petersen, 2021).The emergence of machine learning has facilitated the development of more advanced and innovative technologies, such as generative AI (Hu, 2022).Generative Pre-trained Transformer (GPT) is a model of generative AI that uses a large amount of publicly available digital content data (natural language processing [NLP]) to read and produce human-like text in several languages.GPT demonstrates strong interactive capabilities, being proficient not only in engaging in realistic conversations but also in generating compelling and convincing creative content (Aydin & Karaarslan, 2022;Jovanović & Milosavljević, 2022;Korngiebel & Mooney, 2021;Pavlik, 2023).Generative AI models have achieved a significant breakthrough in on-demand content creation (McKinsey & Company, 2023), particularly in education.AI's various application in education encompasses content development, teaching methodologies, student assessment, and teacher-student communication (Chen et al., 2020;Timms, 2016).
The sophisticated GPT model, ChatGPT, has attracted the attention of various fields, including education (Fütterer et al., 2023;Williams et al., 2023).Within a week of its initial public release on November 30, 2022, ChatGPT quickly amassed over one million subscribers, as reported by Altman (2022) and Mollman (2022).ChatGPT holds significant promise as an asset in the field of education.Educators can leverage ChatGPT to enrich teaching and learning through a multitude of applications, including but not limited to 1) facilitating lesson planning, 2) providing personalised tutoring, 3) automating assignment grading, 4) assisting with language translation, 5) fostering interactive learning experiences, and 6) supporting adaptive learning approaches (Baidoo-Anu & Owusu Ansah, 2023).
The primary objective of this study was to investigate how incorporating AI-generative ChatGPT as a teacher assistant can alleviate teacher workload.A literature review on teacher burnout and ChatGPT explicitly focused on teacher workload and teaching responsibilities.The review identified lesson planning and content development as the most time-consuming and challenging aspects leading to teacher burnout.Expanding upon these discoveries, the testing methodology for ChatGPT was advanced, encompassing the creation of queries that replicate conventional lesson planning and curriculum creation for students aged 12-15 and 16-18, spanning subjects like English, science, and math.The generated lesson plans were evaluated using the 5Es instructional model.
This study is particularly relevant to teacher burnout in the United Arab Emirates (UAE), considering the country's focus on integrating artificial intelligence technologies into the education sector.With the UAE's partnership with Microsoft OpenAI and its commitment to adopting AI tools, such as ChatGPT, in teaching methods (Al Falasi, 2023), the findings of this study can directly contribute to addressing teacher burnout in the country.By exploring how incorporating AI-generative ChatGPT as a teacher assistant can mitigate teacher workload, the study aligns with the UAE's vision of utilising advanced technologies to enhance teaching practices and improve student learning outcomes.The insights gained from this research can inform educational policymakers and stakeholders in the UAE on leveraging AI technologies effectively to reduce teacher burnout and promote teacher job satisfaction and well-being.
The following questions guided the research: 1.
How can using AI generative chatbots, like ChatGPT, mitigate teachers' workload and reduce the risk of burnout?

2.
To what degree can ChatGPT, as an AI generative tool, produce high-quality lesson plans that alleviate teachers' workload and create more time for additional responsibilities?

3.
How can the quality of lesson plans and recommendations for content development produced by ChatGPT be enhanced?

Literature review
To identify factors contributing to teacher burnout and explore how they can be addressed with the assistance of ChatGPT, a literature review was conducted on the topics of teacher burnout and ChatGPT, with a specific focus on teacher workload and non-teaching work time.

Teacher burnout
Burnout has been recognised as a psychological risk, especially in occupations that involve customer interaction (Granziera et al., 2021;Maslach & Leiter, 2016;Montero-Marín et al., 2012).It has been defined as a psychological condition that arises from prolonged exposure to stressors associated with work relationships (Maslach & Leiter, 2016).When work becomes tedious, unfulfilling, and unsatisfactory, burnout develops gradually.
Burnout is typically split into three dimensions: exhaustion, lack of achievement, feelings of scepticism towards the job and disconnection from one's profession.Among the three dimensions, emotional exhaustion is a fundamental aspect of the psychological phenomenon (Maslach & Leiter, 2017).
The adverse effects of teacher burnout are well-documented in the literature and have a negative impact on the well-being of teachers themselves, including their physical health (Hakanen et al., 2006), mental health (Schonfeld & Bianchi, 2016) as well as job satisfaction (Robinson et al., 2019).Burnout has also negatively affected student academic attainment (Herman et al., 2018).Moreover, high levels of burnout are linked to increased absenteeism, attrition, and turnover rates among teachers, as well as a drop in the quality of job performance (Ingersoll & Strong, 2011;Klusmann et al., 2016).Since teacher burnout has wide-ranging consequences for teachers, school systems, and society, it is critical to explore more closely the factors that contribute to burnout among teachers (Granziera et al., 2021).
Several organisational variables were found to be linked to teacher burnout, including the specific demands from teachers, the size of their classes and schools, and the availability of resources and support.A study by Pietarinen et al. (2013) revealed that subject teachers experience significantly higher burnout than class and special education teachers.
Additionally, subject teachers reported the highest levels of inadequacy, while special education teachers reported the lowest levels.The high levels of teacher burnout in subject teachers were reported to be linked to scarcity of resources, workload challenges, and role obscurity (Brunsting et al., 2014).
Prior studies suggest that larger class and school sizes are associated with higher teacher burnout levels.In addition, earlier studies have revealed a correlation between larger school sizes and decreased teacher job satisfaction, reduced levels of job performance, and a sense of demotivation (Pietarinen et al., 2013;Skaalvik & Skaalvik, 2011).Burnout is also associated with teachers' support level, as Brunsting et al. (2014) and Skaalvik and Skaalvik (2011) observed.Teachers who benefited from extra support from trained teaching assistants exhibited lower levels of teacher burnout, mainly when inclusive classrooms accommodated students with special needs (Saloviita & Tolvanen, 2017).
Therefore, it is worthwhile to investigate the role of well-equipped teaching assistants and their potential to lower levels of teacher burnout.
Research across different cultural and geographical settings has highlighted notable elements influencing teachers' workplace well-being (Bermejo-Toro et al., 2016;Dicke et al., 2014;Evers et al., 2016;Skaalvik & Skaalvik, 2011;Tonder & Fourie, 2015).These aspects include excessive workload, student misbehaviour, time constraints, role stress, and students' limited motivation to learn.Studies suggest that these challenges significantly impact teachers' resilience and overall well-being.Heavy workload has been associated with various adverse outcomes for teachers, such as emotional exhaustion, stress, depressive symptoms, lower levels of job commitment, decreased engagement, and a greater likelihood of leaving the profession (Hakanen et al., 2006;Lee, 2019;Leung & Lee, 2006;Skaalvik & Skaalvik, 2011).
According to research, teachers' average time on non-teaching tasks is substantial.Prior teacher workload studies revealed that teachers spend significant time on two non-teaching activities: lesson planning and preparation of teaching and learning materials (Castelli et al., 2017;Salmela-Aro et al., 2019).Findings from a study conducted in England showed that primary teachers self-reported more than 60 working hours per week, a total number of hours that was found to be significantly higher than time spent by secondary teachers (Higton et al., 2017).Secondary school teachers have described increased workload concerning the amount of time spent on class preparation due to the difficulty of a subject they teach (Agyapong et al., 2022).Similarly, teachers from Poland confirmed that extensive non-teaching workload is associated with higher levels of teacher burnout (Mojsa-Kaja et al., 2015).Almost half of the interviewed teachers reported that increased administrative tasks contributed to their already heavy workload (Arvidsson et al., 2019).
The increase in workload was not due to teaching responsibilities but rather the addition of new demands without removing previous tasks.

Artificial intelligence and ChatGPT
Artificial Intelligence (AI) frequently refers to teaching machines that perform tasks that mimic human intelligence.Many everyday technologies, such as chatbots, voice assistants and others, use AI.Machine learning is a subset of AI that involves developing models capable of learning from data patterns without human intervention.The massive amounts of complex data being generated today have increased the potential and necessity for machine learning.Several types of machine learning models build upon classical statistical techniques developed in the past.Until recently, machine learning models were mainly used for predictive purposes, such as identifying patterns in content.An example of a predictive model is Learning Analytics (LA), which combines data mining and artificial intelligence and is often used for predictive and prescriptive purposes.In predictive analytics, LA predicts "what will happen," while in prescriptive analytics, it recommends future actions and interventions (Banihashem et al., 2022).Recent developments in LA enhance study success by collecting and analysing data to provide feedback and support.
Effective techniques detect at-risk students, but there is a need for standardised measures, visualisations, and personalised interventions.However, fully automated systems must balance personalisation and collaboration for effective learning (Ifenthaler & Yau, 2020).
Learning Analytics collects performance data (grades, scores) for educators' early recognition and learning activity data (forum posts, interactions) for student reflection (Banihashem et al., 2022).(Timms, 2016).This has been achieved using different platforms and applications, including Interactive Learning Environments (ILEs), Intelligent Tutoring Systems (ITS), adaptive learning systems, virtual reality (VR) and 3D technology (Chen et al., 2020).In a study by Chassignol et al. (2018), it is highlighted that AI has been used in education administration and teaching, subsequently impacting students' learning.The scholarly sources analysed for the study demonstrate that AI has been utilised in educational institutions in various ways, such as automating administrative tasks, developing curriculum and content, providing instruction, and improving students' learning processes.Through technology such as virtual reality, webbased platforms, robotics, video conferencing, audio-visual files, and 3D technology, students can learn better and have a personalised and richer educational experience.AI has also made teachers more effective and efficient (Chen et al., 2020).Among the opportunities AI provides in education is the further development of massive open online courses (MOOCs), making it easier to assess many assignments and detect learning and teaching gaps.At the same time, these intelligent assessment systems can skip some correct but rare unique solutions since the system decides based on massive statistics.This means that AI-based assessment systems cannot be accurate in every situation without a human mentor (Chassignol et al., 2018).
Generative AI can create new content using unsupervised and semi-supervised machine

Methodology
This section outlines the methodology employed in this study.The researchers utilised an explanatory research design, a research method that explores why something occurs when limited information is available (Ivankova et al., 2006).It can be an assistive tool to increase researchers' perception of a specific topic, check how or why a particular phenomenon occurs, and predict its possible future occurrence (Awotunde et al., 2021).
The ChatGPT testing primarily concentrated on lesson planning and material development for English, science, and math subjects, targeting students aged 12-15 and 16-18.The lesson plans generated through ChatGPT were evaluated according to the 5Es teaching model (see Table 2).Researchers used a set of ChatGPT prompts during this testing process to create lesson plans for specific subjects and age groups, as described earlier.The main focus of the research is to examine ChatGPT's potential in designing effective and engaging learning materials.This approach did not include direct interaction with students but helped judge the tool's proficiency in generating lesson plans mapped to the 5Es instructional pedagogical model aligned with the constructivism theory (Chinwe & Chinyere, 2010).This exploration into ChatGPT's capabilities sheds light on its potential to significantly assist teachers in developing engaging and comprehensive learning content for diverse students.

ChatGPT testing
Two primary criteria shaped and guided the content and scope of ChatGPT testing.Firstly, the literature review findings identified lesson planning and content development as the most time-consuming non-teaching tasks contributing to teachers' workload and burnout.
These tasks require teachers to carefully consider learners' distinct needs, learning objectives, and appropriate resources and activities, which can significantly impact their overall workload and well-being.Therefore, teachers must find ways to streamline these processes to reduce stress and manage workload.Secondly, subject teachers have reported the highest levels of inadequacy due to workload problems, ambiguity, and lack of resources.Based on these two findings, the ChatGPT testing for this study focused on lesson planning and material development cycles two and three, covering English, science, and math for 12-15 and 16-18-year-old students.The testing queries were designed to simulate everyday tasks of lesson planning and material intended for classroom teaching.

Data analysis of the ChatGPT-generated lesson plans
To determine its effectiveness, the lesson plans generated by ChatGPT were analysed and evaluated using the 5Es instructional model for learning and teaching to assess if it can create quality lesson plans that help teachers reduce their workload and free up some of their time for other tasks.The first and the second authors evaluated the lesson plans generated by ChatGPT using a checklist of the various descriptors of each of the 5E phases (see Table 2).The authors reviewed the lesson plans, and based on the 5E descriptors, they evaluated each lesson plan and recorded qualitative feedback regarding commendable aspects and areas for improvement.
The underlying dynamic and interactive learning theory justifies using the 5Es instructional model as an evaluation framework for lesson plans developed by ChatGPT.Individuals revise their initial concepts through interactions with the environment and others, redefining and reorganising their understanding.The learner interprets objects and phenomena based on current experiences and incorporates these interpretations into their knowledge.The 5E Instructional Model, or the 5Es, consists of the following phases: engagement, exploration, explanation, elaboration, and evaluation.Each of the 5E phases is defined in Table 2.The model acknowledges that students construct knowledge through reconstruction, which can be facilitated through well-designed lesson sequences that challenge their existing conceptions (Bybee, 2006).

Summary Engage
The teacher stimulates students' prior knowledge and involves them in activities that promote current concepts.These activities should link prior knowledge with the desired learning outcomes to organise learners' thinking.

Explore
The students conduct activities that detect misconceptions or promote conceptual change.Students employ questioning skills and brainstorming to reveal new insights and ideas to help them investigate.

Explain
Students pay attention to a specific learning experience, demonstrating their knowledge and using what they learned.Students may explain new concepts with structured and guided support from teachers and the curriculum.Elaborate At this stage, students are challenged to elaborate on their conceptual understanding; they are required to expand the learning experience to other learning contexts, which leads to a deeper perception of the concepts' core ideas.Evaluate The teachers encourage students to evaluate their understanding and their application of learning experiences, where teachers can monitor student's progress in acquiring new knowledge and skills about the concepts.

Findings of ChatGPT testing
The ChatGPT testing aimed to investigate its effectiveness in generating lesson plans and teaching materials as a potential solution to reduce the time required for these tasks.The lesson plans and adaptations that ChatGPT developed in response to our testing prompts were evaluated based on the 5Es Instructional Model, and our evaluation yielded the findings illustrated in the following section.diverse assessment strategies such as performance tasks and projects, could benefit the lesson plan.In summary, the findings suggest that while the lesson plan displays several strengths, several areas require refinement to enhance its overall efficacy.The results are summarised in Table 3.The introduction could be more engaging and interactive.Explore A variety of activities that allow students to explore the concepts of square roots and cube roots.

ChatGPT Test
Could benefit from more interactive activities and strategies for addressing misconceptions.

Explain
Clear explanations of concepts and opportunities for students to create their problems.
Could benefit from more visual aids and real-world examples.

Elaborate
Challenging activities that require students to apply concepts and work in groups.
This could include more opportunities to apply concepts to real-world problems and more challenging problems.Evaluate Short quiz at the end of class and encouragement for students to ask questions and share insights.
Could benefit from more ongoing assessment and feedback using a variety of strategies.Incorporates a variety of assessment strategies to evaluate students' grasp of the concepts and provides continuous feedback and assistance to aid them in enhancing their comprehension.

The evaluation indicates that
evaluate their comprehension while providing continuous feedback and support to enhance their learning.In contrast to the original version, the modified lesson plan maintains greater comprehensiveness and incorporates more learner-engaging elements.Additionally, the adjusted version is better suited to effectively facilitate students' understanding and proficiency in grasping square roots and cube roots.These results are summarised in Table 4.
ChatGPT can also develop the content and propose the supplementary material needed to accomplish the activities.For the lesson mentioned above plan, we requested ChatGPT to provide examples of real-life problems, and it did.

Prompt Three:
Provide examples of real-life problems for the evaluation step.
The prompts were replicated with two more subjects to decide if ChatGPT can generate accurate lesson plans with multiple subjects and grade levels.ChatGPT opted to create a chemistry lesson plan using the 5Es model without being prompted.This demonstrates the model's adaptability to the given input.We asked ChatGPT to explain his answer in the following prompt.

ChatGPT Test
Prompt Two: Can you explain why you used the 5Es framework for the lesson plan you designed when the first plan you created did not follow that structure?Here is the answer: We decided to increase the complexity of the task assigned to ChatGPT to assess its proficiency in generating lesson plans that adhere to the 5Es framework while also incorporating differentiation to cater to the diverse needs of learners.
Prompt Three: Could you modify the above lesson plan to accommodate students of different levels, including lower achievers, middle achievers, and high achievers?Here is the answer: Regarding content development, we asked ChatGPT to generate a formative assessment for the evaluation stage.
Here is the answer: ChatGPT Test Three: A lesson plan for Grade 7 English (Creative Writing) Prompt One: I am teaching creative writing to students in Grade 7. Our topic is a story about an ant struggling not to be washed away by the rain.The aim is to teach students the value of perseverance to achieve success.Create a lesson plan within a 45-minute class period.
This time, we asked ChatGPT to evaluate the lesson plan it has created using the 5Es framework to check if it aligns with the 5Es framework and effectively engages students in the learning process.
Prompt Two: Use the 5Es framework to evaluate the effectiveness of the above-generated lesson plan.
Subsequently, ChatGPT was tasked with utilising its evaluation to improve the lesson plan it had produced under the 5Es framework.
Prompt Three: Can you improve the lesson plan based on your provided evaluation?
The evaluations conducted on ChatGPT revealed its ability to generate lesson plans that can be improved.ChatGPT demonstrated its capability to utilise user prompts and produce responses more tailored to the user's needs, as exemplified by the chemistry lesson plan illustration.Interestingly, ChatGPT automatically employed the 5E's framework to design the chemistry lesson plan without specific instructions, likely because of the previous request to create a math lesson plan using the 5Es.However, ChatGPT did not recognise the need to apply the framework when the prompts were not made simultaneously.This was evident when ChatGPT did not use the 5E's to develop the English lesson plan in the first trial, which can be attributed to the fact that the English Lesson plan was developed independently from the other two tests that focused on math and chemistry.

Discussion
The current study aimed to investigate how using AI generative ChatGPT as a teacher assistant can help reduce teachers' workload and mitigate the risk of burnout.A review of the current body of literature revealed that significant demands characterise the teaching profession and is a prominent contributor to teacher burnout.This can be attributed to many factors, including but not limited to excessive workload, insufficient support, behavioural challenges exhibited by students, challenges in managing time effectively, and inadequate access to necessary resources (Bas, 2011;Denny et al., 2011;Diehl & Carlotto, 2014;Fernet et al., 2012;Shernoff et al., 2011).Teacher burnout has significantly impacted the attrition rates of experienced teachers, resulting in a shortage of qualified teachers (Fernet et al., 2012;Ingersoll & Strong, 2011;Skinner & Jeffry, 2016).Therefore, it is critical to acknowledge and tackle the issue of teacher burnout to enhance teachers' well-being and guarantee high-quality education (Cozolino, 2017;Kaschka et al., 2011).
Through a comprehensive review of the literature, this study has revealed that teacher burnout is predominately influenced by two key factors: heavy workload and inadequate support (Brunsting et al., 2014;Hakanen et al., 2006;Lee, 2019;Leung & Lee, 2006;Skaalvik & Skaalvik, 2011).Previous studies have shown that these factors have a significant role in contributing to teacher burnout.Resolving these factors through supportive practices is crucial to improve teacher well-being, minimise workload, and prevent burnout.One notable approach is using AI technologies, such as ChatGPT, to function as teaching assistants and help teachers perform tasks that impose a substantial workload.Several research studies indicate that ChatGPT can support teachers differently (Farrokhnia et al., 2023).Teachers can use ChatGPT as a virtual assistant to receive quick and reliable answers to inquiries related to pedagogical approaches, subject-specific content, lesson planning, and classroom management strategies (Cotton et al., 2023;George et al., 2023;Mizumoto & Eguchi, 2023;Qadir, 2022;Zhai, 2022).
The results obtained from the ChatGPT testing align with the findings documented in the literature review.They provide evidence that ChatGPT has the potential to support teachers and mitigate specific challenges they encounter, ultimately resulting in an enhancement in their work-life balance (Kilinç, 2023).The results suggest that ChatGPT can provide personalised recommendations for lesson planning and preparation of teaching and learning materials.Providing teachers with a lesson plan generation tool allows them to allocate more time and attention to other salient facets of their profession.Empirical findings stemming from ChatGPT's testing in creating lesson plans for various subjects reveal that through the input of information on curriculum, learning objectives, learning theories, instructional models, and student requirements, ChatGPT can swiftly generate lesson plans and educational materials of high quality that are customised to address the particularities of each class and student which is also confirmed by previous studies (Farrokhnia et al., 2023;Kasneci et al., 2023;Qadir, 2022;Rahman & Watanobe, 2023).
Teachers can use ChatGPT to streamline the planning process, enabling them to devote their time to other important aspects of teaching, including classroom instruction and student support.Previous research on using artificial intelligence in education supports this finding by providing evidence that AI technology can provide teachers with practical tools to help them manage their workload (Papamitsiou & Economides, 2014).ChatGPT also allows teachers to provide their school's lesson planning templates so that ChatGPT can generate lesson plans based on the template.This was tested and proved successful in the current study when ChatGPT was asked to create a lesson plan using the 5Es template.The generated lesson plan can be customised by teachers based on their schools and student's specific requirements and needs.This was supported by the research findings that stressed the importance of customisation (providing personalised instruction) in improving students' learning (Kirschner & Merriënboer, 2013).
However, it is noted that the quality of the lesson plans generated by ChatGPT was not achieved on the first attempt.It was completed after providing ChatGPT with many specific prompts.The reasons behind this might be that ChatGPT lacks a deep understanding of the questions asked, and its comprehension is limited, especially for tasks like lesson planning that necessitate a comprehensive understanding of domain-specific knowledge and context (Bogost, 2022;Borji, 2023;Gao et al., 2023;Gupta et al., 2023).
These results are supported by an empirical study conducted by Sinha et al. (2023), who argued that ChatGPT lacked a deep understanding of the theoretical concepts, affecting its ability to generate quality responses to complex problems.The present research addressed this limitation by furnishing ChatGPT with detailed instructions that assisted it in acquiring a more profound and less surface-level comprehension, enhancing the meaningfulness of its prompts.However, if "such a process is lacking, the use of ChatGPT could lead to a kind of 'dumbing down' of the users" (Farrokhnia et al., 2023, p. 7).
This study's results indicate that prompts significantly affect the effectiveness of generated lesson plans by ChatGPT.Specific and explicit instructions added to the AI model increase the lessons' relevancy and usefulness.A crucial component of using ChatGPT as a teacher's assistant is using thoughtfully designed and task-specific prompts.
Through this approach, ChatGPT offers personalised recommendations tailored to each teacher's unique needs and challenges (Rudolph et al., 2023;Shen et al., 2023).ChatGPT can also produce unique responses in different tones and patterns to meet the needs of various teachers (Aljanabi, 2023).
Therefore, teachers are recommended to exercise caution when accepting the initial response (lesson plan) proposed by ChatGPT.Rather than implementing a proposed plan immediately, it is imperative to evaluate it comprehensively in terms of the criteria specified, the needs of students, and any other requirements.Likewise, to improve ChatGPT's accuracy and relevance, teachers should engage in a feedback loop with ChatGPT.This involves providing input to ChatGPT, then creating more precise prompts based on ChatGPT's responses.Moreover, teachers must provide ChatGPT with clear, accurate prompts tailored to their unique needs and pedagogical approach (Gilson et al., 2023).
Another significant finding from the current study is the potential of ChatGPT to learn and change its responses based on its engagement and interactions with human users (Shen et al., 2023).The current study also found that ChatGPT could learn and alter its responses based on interaction and engagement with human users (Shen et al., 2023).ChatGPT demonstrated the ability to recall and integrate previous interactions into its responses when it was asked to generate a lesson plan without explicitly referencing the 5Es framework in the subsequent instance.As a result of this adaptive capability, ChatGPT can retain contextual information over time and maintain a cohesive conversation with users.

Implications for research and practice
The current study focused only on investigating and evaluating the potential benefits of AI generative ChatGPT in addressing teacher burnout.Therefore, further research is needed to understand the impact of AI integration in education and its influence on teaching and learning.More research is required to determine the effectiveness of AI as a teacher assistant and its effect on students' results.
In the context of the UAE, where the integration of AI technology in the education sector is of top priority, the study's findings provide important implications for policymakers and stakeholders.Due to UAE's partnership with Microsoft OpenAI and its goal to integrate AI technology (e.g., ChatGPT) in teaching and learning (Al Falasi, 2023), the results of this study are highly relevant to the problem of teacher burnout in the region.This study's results align with the UAE's goal of using advanced technology to improve teaching practices and students' results.This study's findings can guide educational policymakers and stakeholders in the UAE to determine the best ways to integrate AI technologies to reduce teacher burnout and enhance teacher job satisfaction and well-being.
The findings of this study showed that ChatGPT recommendations can assist teachers in developing personalised lesson plans, content, and student plans.AI technologies can improve education quality by empowering teachers with tools that make their daily tasks easier to manage.Schools should consider using artificial intelligence (AI) tools (such as ChatGPT) to reduce teacher workload and burnout.In addition, teachers need to balance the use of artificial intelligence and their human knowledge and input to the produced materials by ChatGPT.
The results of the current study also indicate that prompts play a vital role in using ChatGPT to produce effective lesson plans.This finding suggests that teachers need to use clear and precise prompts that maximise the benefits of ChatGPT as their assistant.
Therefore, academic institutions should provide practical workshops to train teachers on using ChatGPT properly and how teachers can design clear and concise prompts that target specific activities.

Conclusion
This research aimed at studying the efficacy of AI generative ChatGPT as a teaching assistant to mitigate teacher workload and avert burnout.Based on the literature review, much critical evidence supports the claim that teaching is demanding and can cause burnout due to various tasks and requirements and the continuously stretchable profession's quality standards.This research aligns with previous findings confirming the positive and significant relationship between workload and teacher burnout.
Regarding generative AI models, they can now create new content on demand, a significant breakthrough (McKinsey & Company, 2023).The application of AI in education provides an opportunity to break physical barriers as learning materials are now accessible online.Different studies have demonstrated the transition of AI in education from conventional computers to embedded systems such as robots.AI in education goes beyond the normal functions of computers and involves the collaboration of different professionals, including data scientists, product designers, linguists, cognitive scientists, psychologists, and education experts.AI surpasses the conventional understanding of various technological applications in education.AI in education has been applied in other areas, such as content development, teaching methods, student assessment, and communication between teachers and students One: A lesson plan for Grade 8 Math Prompt One: Can you design a lesson plan for an eighth-grade class focusing on square roots and cube roots within a 45-minute class period?Evaluation of the lesson plan using the 5Es instructional design framework: The assessment of the lesson plan revealed commendable elements, including a well-defined introduction, a diverse range of activities for concept exploration, lucid explanations, thought-provoking tasks, and a concise quiz for evaluation.Nonetheless, there exists room for improvement regarding the introduction's level of engagement and the incorporation of interactive and visual aids in the plan.Strategies for correcting misconceptions are absent from the lesson plan, and opportunities for real-world problem-solving ought to be incorporated.The addition of continuous assessment and feedback mechanisms, including ChatGPT failed with the first trial in creating a lesson plan that adheres to the 5Es framework and the constructivist theory expectations.Therefore, a new prompt was fed into the chatbot, requesting it to make the necessary changes to incorporate the needed improvements.Prompt Two: Can you fix the lesson plan above, incorporating the improvements needed to match the 5Es?The updated lesson plan, which adheres to the principles of the 5Es model, demonstrates a more student-centred and interactive approach.The lesson plan incorporates discoverybased activities and group work, as well as visual aids and real-world examples to aid students in grasping the concepts.Additionally, it offers more opportunities for students to apply these concepts to real-world problems.It employs various assessment strategies to Two: A lesson plan for Grade 10 Science (Chemistry) Prompt One: Can you design a lesson plan for a tenth-grade class focusing on Decomposition Reaction within a 45-minute class period?
The study revealed ChatGPT's potential to support teachers in various tasks covered in previous research.It highlighted AI's role in improving teaching and learning practices, thus relieving teachers of burdensome major time-consuming tasks.By getting effective planning support, teachers can allocate more time to improve teaching practices, assessment for learning, and student engagement.Although ChatGPT's initial answers might lack depth, careful, precise, and detailed instruction can enhance comprehensibility, reducing the risk of oversimplification.Skilfully created prompts prove the efficiency and quality of generated responses, showcasing the importance of collaboration between AI and human expertise to retain the context of the task given to ChatGPT while processing it.In other words, the study emphasises the importance of balancing ChatGPT and teachers' input, acknowledging teachers' indispensable and substantial roles in integrating AI into lesson planning and preventing burnout.There are many implications for research, and further exploration into AI's impact on the teaching profession and its required tasks, responsibilities and student achievement is warranted.As UAE encourages and promotes AI integration in learning and teaching, stakeholders and policymakers can use the findings to feed into potential educational reform movements, as the study highlights ChatGPT's potential for designing personalised and customised learning experiences and effective lesson planning.Educational institutions should consider ChatGPT as an assistive tool, which requires them to provide proper training on creating generative prompts for quality responses in light of ethical use.The careful and ethical use of ChatGPT as a supportive tool promotes excellence in education.Limitations and directions for future researchThis research includes certain limitations that should be considered.The highlighted ChatGPT testing in this research might not entirely reflect the complex real classrooms, affecting the ability to generalise the results.While the study focused on lesson planning and teaching content development, teaching includes many responsibilities, which this research did not cover holistically.Qualitative analysis and subjective evaluation of ChatGPT-generated content could result in a variable quality assessment reflecting different teaching standards.Furthermore, the study focuses on ChatGPT, one type or tool from a broader range of AI models.Integrating AI tools such as ChatGPT in education should comply with ethical use, including data privacy and data biases, as Abu Khurma et al. (2023) highlighted that ChatGPT depends on the data it was trained on, which requires further exploration.In summary, while discussing insights about the potential benefits of the AI tool (ChatGPT), this study acknowledges the need for further research to refine AI's tool (ChatGPT) role in the teaching profession while addressing these limitations.In addition, it is beneficial to conduct more research on ChatGPT potential to improve motivation and drop-out prevention from online professional development platforms, such as and not limited to MOOCs, where ChatGPT could address the drop-out discussed byBadali et al. (2022) and related to the following motivational types: "Fulfillment of the need for competence, usefulness, personal growth, etc.", as ChatGPT has a potential of streamlining the completion requirements of these courses and detecting possible learning gaps for teachers as professional learners.
Answering inquiries: ChatGPT can address a range of inquiries, including academics, curricula, lesson preparation, and instructional techniques.
(McKinsey & Company, 2023)as text, audio, video, images, and code(McKinsey & Company, 2023).Such technologies include Generative Adversarial Networks (GANs) and Transformer-based models like Generative Pre-trained Transformer (GPT).However, generative AI must be modelled to enable it to create new content from existing data.Creating a generative AI model has traditionally been a daunting task, typically undertaken by only a handful of tech giants with extensive resources.Well-known donors have provided billions in funding to companies like OpenAI, which has produced ChatGPT, previous GPT models, and DALL-E.DeepMind, a subsidiary of Alphabet, Google's parent company, and Meta, have also released products that leverage generative AI.These businesses employ some of the world's most accomplished computer scientists and engineers, but training a model with vast internet data incurs significant expenses.While OpenAI has not revealed precise figures, it is believed that GPT-3 was trained on approximately 45 terabytes of text data at the cost of several million dollars.ChatGPT, developed by OpenAI, is a free chatbot that can generate answers to questions and has already gained a million users.Despite fears about its impact, machine learning has positively affected various industries.The adoption and investment in AI have significantly increased in the past five years.However, the full extent of the impact and risks of generative AI is still unknown(McKinsey & Company, 2023).AI chatbots have previously received poor reviews, but ChatGPT appears to perform well despite still being conclusion of a phrase from just a few words.The success of ChatGPT demonstrates that these models can be highly accurate with sufficient example text.In its capacity as an AI language model, ChatGPT may assist teachers with a range of daily duties, including:   Providing language translation: ChatGPT can translate the language for teachers who work with multilingual students or need to communicate with parents who speak different languages.Integrating AI principles and technology into web-based learning platforms, such as Adaptive and Intelligent Web-Based Educational Systems (AIWBES) and Intelligent Web-Based Education (IWBE), improves the learning experience by considering several factors such as the learner's knowledge, skill, and performance capabilities.Findings from the literature review are presented in Table1.

Table 1
Summary of the literature review findings

Table 3
Summary of the evaluation of the lesson plan

Table 4
Comparison between the two lesson plans created by ChatGPT