Technology Review of Magic School AI: An Intelligent Way for Education Inclusivity and Teacher Workload Reduction

Abstract

Students with special needs often require more assistance and attention to meet their educational needs. However, schools frequently grapple with a critical shortage of special education teachers and support staff. This shortage of special education teachers can result in limited resources for general and subject teachers (e.g., math, science), making it challenging to provide individualized support to students with special needs. Specifically, subject teachers may struggle to design effective curricular content modifications and accommodations for such students without the guidance and suggestions of special education teachers. Artificial Intelligence (AI) technologies can provide some support for teachers and schools in meeting the needs of students with special needs. Also, AI may help reduce teachers’ workload. In this technology review, we assess the capabilities of Magic School AI (MSAI) in providing accommodations and modifications to assist teachers in streamlining their workload and fostering inclusivity in their classrooms. We examined five functions: text leveler, text scaffolders, assignment scaffolder, exemplar and non-examples, and sentence starters. Additionally, we discuss the limitations of MSAI and conclude by suggesting potential improvements for the system.

1. Product at a Glance

The details of the technology can be found in

Table 1. Magic School AI Feature.

Product Name	Magic School AI
Website	https://www.magicschool.ai/ (accessed on 17 June 2025)
Languages	English
Product Type	A web-based Artificial Intelligence Platform for teachers
Operating System	PC or Mac with Chrome or Safari
Hardware Requirements	A device can connect to the Internet
Price	Free

.

2. Introduction

Education equity and equality, particularly in the context of special education, remain a critical concern for schools, teachers, and parents. The pursuit of educational equity strives to ensure all students, especially those with Individual Education Plans (IEPs) due to disabilities such as learning disabilities and concentration deficits, receive the necessary resources (McCabe & Ruppar, 2023). The range of disabilities students may have is extensive. The range of disabilities includes 13 categories: specific learning disability (SLD); other health impairment (OHI); Autism spectrum disorder (ASD); emotional disturbance; speech or language impairment; visual impairment; hearing impairment; deafness; deaf-blindness; orthopedic impairment; intellectual disability; traumatic brain injury; and multiple disabilities. These varied needs necessitate collaboration between special and general education teachers or subject teachers to support the development of tailored learning objectives and approaches for students with IEPs.

A major challenge in achieving this collaboration is the understaffing prevalent in schools, significantly affecting special education (Jameson & Huefner, 2006; Nichols et al., 2008; The Children’s Guild, 2024). Moreover, 40% of special education teachers leave the field within the first five years, which results in a high turnover rate (Jameson & Huefner, 2006). In addition, special education teachers, specialists in their field of education, are vital for supporting and co-teaching with colleagues in core subjects like English Literature, Math, and Science (Vannest & Hagan-Burke, 2010). However, they often find themselves overextended (Eisenman et al., 2011; Morgan, 2016). This overextension leads to insufficient support for teachers, especially those of non-core subjects (e.g., foreign languages, music, PE, etc.), resulting in increased workload and stress for general and subject teachers and less inclusive classrooms. Teachers without specialized training often fail to provide appropriate accommodations and modifications (Chu et al., 2020; Cutanda-López & Alfageme-González, 2022), potentially hindering students’ academic engagement and growth (Gesel et al., 2021). Recognizing these challenges underscores the need for broader teacher training in special education and a more equitable distribution of resources across all subjects (Ballhysa & Flagler, 2011; Cooc, 2019).

Amid these challenges, emerging technologies offer innovative solutions. Carr (2023) indicated that Artificial Intelligence (AI) could support teachers to create inclusive lessons by tailoring the content to address the learning gap. In addition, Petera (2023) demonstrates that AI can recognize students’ potential misconceptions and provide teachers with an alternative instructional approach. However, specific applications in the field of special education have not been thoroughly explored.

Magic School AI (MSAI) is a breakthrough AI tool that can be applied to special education. Both general teachers and special education teachers can benefit from using it. It can provide modification and accommodation suggestions for general teachers to promote an inclusive teaching and learning environment. Besides, it can reduce the workload of special education teachers and directors by generating comprehensive reports.

In addition, MSAI is an affordable support tool for subject teachers who do not receive adequate special education resources. The researchers chose MSAI because of its practical value in aiding general education teachers in effectively providing appropriate support for special education students in their students, especially in content areas such as science, social studies, arts, PE, etc., who may lack both previous formal special education training (Begeske et al., 2023) or ongoing structured teaching support (Özgüç & Cavkaytar, 2015). As a free and easy-to-operate platform, MSAI requires minimal training for practical usage.

The features of MSAI are intuitively organized within an accessible portal, allowing teachers to quickly locate the correct tools based on their teaching goals and meet diverse student needs through lesson planning, differentiation, and individualized content creation. The evaluation focused on effectively using the app to bridge the gap by developing differentiated teaching materials, adapting accessible class instructions, and brainstorming inclusive teaching strategies.

Therefore, in this technology review, the researchers critically examine how MSAI can aid teachers in providing modifications and accommodations so that students with IEPs can access instructional content and be included in the classroom. The researchers also assess how MSAI can help special education teachers and directors reduce workload. Our assessment focuses on the tool’s effectiveness and practicality, considering broader systemic changes necessary for addressing the overarching challenges in special education.

3. Literature Review

The lesson planning process is a critical aspect of educational pedagogy, serving as a detailed guide for educators to outline objectives, activities, and expected outcomes for each lesson (Hall, 2019). Lesson planning is a key tool in ensuring a well-balanced and time-managed approach to teaching and facilitates the effective use of resources, including technology. Lesson plans also function as a communicative tool, helping teachers address various teaching challenges and assumptions (Janssen & Lazonder, 2015; Y. Lee & Takahashi, 2011). An essential component of lesson planning is considering the diverse needs of all students, including those with Individual Education Plans (IEPs), and integrating appropriate accommodations and modifications.

Harrison et al. (2013) provide a helpful set of definitions for the types of supports that an IEP may include: (1) modifications, which are “practices that alter expectations to compensate for a disability”; (2) accommodations, which are “changes that maintain the expectations for the child, but provide a differential support to help in the meeting of the standards”; and (3) interventions, “changes made through a systematic process to develop or improve knowledge, skills, behaviors, cognitions, or emotions” (p. 556).

Accommodations are supportive measures provided to students, especially those with special needs, to ensure equal access to the curriculum without altering its content or expected outcomes. Examples include extended test times and the use of assistive technology. To aid teachers in conceptualizing suitable accommodations, models such as preexisting lesson plans are recommended (Martínez & Porter, 2018). On the other hand, modifications refer to changes made to the curriculum or assessment methods to assist students with special needs, potentially altering assignment content, complexity, or expectations. These modifications can assist in ensuring that educational content is accessible and beneficial to all students (Lim et al., 2018).

Teachers, however, often encounter difficulties in providing these accommodations and modifications for these students due to limited access to instructional materials, resources, and support (Cuñado & Abocejo, 2019). Art teachers interviewed reported having limited professional development opportunities to collaborate with special education personnel (Burdick & Causton-Theoharis, 2012; Guay, 2003). Students with special needs constantly meet challenges in reading and writing, which limit their academic development and, therefore, cause lower performance in other subjects. Villanueva et al.’s (2012) comprehensive meta-analysis showed that inquiry-based instruction, especially when paired with appropriate scaffolding and support, can significantly enhance science achievement for these students. Yet, a four-month study observing a middle school science and technology class that included students with mild intellectual disabilities also revealed a contradiction between the need for differentiated and adapted instruction and the teacher’s lack of resources to enhance teaching effectiveness (Özgüç & Cavkaytar, 2015).

In this context, integrating Artificial Intelligence (AI) into education, particularly in providing accommodations and modifications, emerges as a promising solution. AI has the potential to streamline the lesson planning process by leveraging data analysis to offer tailored recommendations for individual student needs (Paiz et al., 2025; Nguyen et al., 2023). Research showed that preservice teachers who incorporated AI-driven modifications into pre-existing lesson plans produced more effective and inclusive lesson plans while gaining increased confidence in their teaching abilities (Lim et al., 2018; Kehoe, 2023; Acquah et al., 2024). Lampou (2023) and Mollick and Mollick (2023) also show that AI infused in lesson planning can make learning more accessible by providing multiple explanations, solving students’ misconceptions, and offering additional scaffolding to students with special needs and language barriers.

AI’s application in accommodations and modifications extends to providing personalized learning experiences and adapting content to individual student needs (Amzil et al., 2023). It also offers predictive analytics to assist educators in making informed instructional decisions. However, despite these advances, there are still significant research gaps in exploring the full potential of AI in aiding teachers in providing modifications and accommodations.

4. Results

4.1. Potential Benefits and Implications of Magic School AI

Magic School AI (MSAI) represents a transformative tool for teachers, simplifying the design of lesson plans, rubrics, assignments, assessments, and even editing class-appropriate videos. Its diverse functionalities promise a more inclusive class by providing accommodations and modifications. In this technology review, we focus on the functions of subject teachers to include students with special needs, which is under the student support category. These functions include Exemplar and Non-exemplar, Text Levelers, Assignment Scaffolders, Text Scaffolders, and Sentence Starters to evaluate and understand how MSAI facilitates the creation of inclusive learning and enhances teaching effectiveness. We categorize all the functions through three perspectives: support modification, assignment scaffolder, and workload reduction.

4.2. Support Modification

Text Levelers

Modification aims to modify the task and make it aligned with students’ levels. Students with ADHD or reading disabilities cannot read for long due to their attention deficit (Toplak et al., 2003). Most special education students’ reading levels are also behind their counterparts (Lewandowski et al., 2015). Teachers can use text levelers to adapt the text to their reading levels or grade levels.

In Figure 1, teachers can upload any documents to the MSAI and choose the appropriate levels they want the text to be. I used a text from the CommonLit Website. This is a 9th-grade level with Lexile: 1140 news.

In Figure 2, we chose 7th grade as the reading level, which summarizes the main idea of stem cells’ benefits and potential applications. This summary helps special education students to gain an understanding of the reading quickly. Besides, the content area, language, and vocabulary are accessible to 7th-grade students. In Figure 3, we changed 7th to 11th grade with the exact text. The output is different. This text uses more scientific, academic vocabulary and more complex sentences. Teachers can use this function to modify the text for special education students in their class without spending too much extra time. If a teacher taught a 9th-grade class with different reading levels, such as 6th grade, 7th grade, and 9th grade, the teacher can simply differentiate the text levels by simply selecting the grade level in MSAI.

4.3. Support Scaffolding and Access

Assignment Scaffolder

Students who have IEPs need scaffolding and support for their assignments. Teachers can use the assignment scaffolder function, which will be broken down into multiple steps and made manageable and understandable for students. This not only reduces teachers’ workload through generating step-by-step scaffolding but also benefits students who need assistance to process the multiple-step task.

Figure 4 shows the assignment prompt, which is long and contains multiple questions. We chose the 9th grade as the grade level to ask MSAI scaffolding questions for students.

Figure 5 indicates that MSAI breaks down the task into seven steps. Each step starts with clearly telling the audience what to do, with two bullet points. For example, MSAI asked the students to “write down their impression of ‘story-teller’ in each work” in Step 4. This visualization and simplification essentially help special education students to process the information effectively. Afterward, MSAI continued to deepen students’ understanding by prompting them to compare the similarities and differences in how they convey stories.

4.4. Support Accommodation

4.4.1. Sentence Starters

Sentence starters as an effective accommodation can scaffold students with special needs. Figure 6 shows the instructions for 11th-grade students to describe the phases of mitosis. Students can use the sentence starters to describe mitosis phases (Figure 7).

Moreover, teachers can use the ChatGPT 3.0-like function to ask follow-up questions, which supports scaffolding and engagement. In Figure 8, we asked it to provide a visual representation of mitosis phases. This visualization helps students with special needs to get a precise flow of mitosis phases.

4.4.2. Text Scaffolder

Most special education students may be behind grade level compared to their counterparts. Text scaffolders modify the complexity and length of the text. This enables students to get access to the content. In addition, this function provides teachers with some level-appropriate questions to check students’ understanding.

Researchers adopted a time machine story (ReadingVine, 2022) as an example. This text is appropriate for the 7th-grade level and has 451 words. Researchers imputed the grade level, the vocabulary that they wanted to define, and the questions they wanted to ask (Figure 9). The maximum vocabulary that can be defined is 7, and the maximum number of literal questions that can be generated is 7. Researchers chose seven vocabulary words to define and three literal questions to ask (Figure 10). It can accurately define the Tier 2 and Tier 3 vocabulary words. It also provides three questions for teachers to assess readers’ comprehension.

4.5. Supports Access and Comprehension

4.5.1. Exemplar & Non-Exemplar

The exemplar and non-exemplar functions provided by MSAI serve as valuable tools for teachers working with special education students. The exemplar sets clear expectations and serves as a model for special education students to finish their assignments effectively. In addition, all students, regardless of their educational needs, can benefit from the clarity provided by exemplars and non-exemplars as they embark on their assignments.

To illustrate, we chose a Common Core State Standard (CCSS) for 9th grade: ELA.RL.9-10.1: Cite strong and thorough textual evidence to support explicit analysis of what the text says and inferences drawn from the text. The text the researchers chose was To Kill a Monckingbird (H. Lee, 2010, p. 119). By specifying the grade level, CCSS standards, length, and prompt (as shown in Figure 11), MSAI generated an exemplar and a non-example (as depicted in Figure 12). Besides, it also provides explanations for the errors made by non-exemplars.

4.5.2. Follow-Up Question

MSAI allows users to ask follow-up questions to better understand the content. There are two ways users can do this. First, MSAI provides suggested follow-up questions as clickable prompts. Alternatively, users can type in their own questions—for example, asking MSAI to suggest more hands-on activities.

In Figure 13, a 9th-grade reading passage titled Summer Rain was uploaded to MSAI’s text scaffolder. The user asked MSAI to define seven vocabulary words and generate five literal comprehension questions. The results are shown in Figure 14. Below the output, MSAI suggested two follow-up questions. The researchers selected the prompt, “Could you inquire how the narrator’s feelings about summer rain evolve over time?” as an example. In Figure 15, MSAI responded with five new questions related to this prompt.

The researchers then continued by entering their follow-up question: “Can you provide some visual aids for this reading?”—without using a suggested prompt. As shown in Figure 16, MSAI responded with five potential visual aids: a diagram, a timeline, an emotion chart, an illustrative image, and a graphic organizer. In addition, MSAI provided detailed descriptions of what each visual aid should include, significantly reducing teachers’ workload.

5. Summary of MSAI Features

The details of the MSAI features and functions can be found in

Table 2. Summary of MSAI Features and Function.

MSAI Feature	Pedagogical Function	Modification/Accommodation/Workload Reduction	Explanation
Text Leveler	Adjusts reading difficulty based on grade level	Modification Workload Reduction	Changes text complexity to fit students’ reading levels; saves teachers time rewriting materials.
Text Scaffolder	Defines vocabulary, generates guiding questions, shortens or explains texts	Accommodation Workload Reduction	Supports access to grade-level text without changing standards and is helpful for students with IEPs or ELLs.
Assignment Scaffolder	Breaks down multi-step assignments into manageable steps	Accommodation Workload Reduction	Helps students with executive functioning challenges follow complex tasks; saves planning time.
Exemplars & Non-Exemplars	Models strong and weak student work with explanations	Accommodation Workload Reduction	Clarifies expectations and reduces ambiguity for students who benefit from concrete examples.
Sentence Starters	Provides structured sentence stems for writing and academic tasks	Accommodation	Aids students who struggle with language production or organization.
Follow-up Questions	This is for an additional function for all features in MSAI	Workload Reduction	Individualized the content based on teachers needs

.

6. Limitations and Implications

While MSAI offers significant advantages, it’s essential to consider its limitations. A key issue is the lack of adequate technological training for teachers, leading to varying degrees of acceptance and understanding of new technologies. This disparity can impact their ability to effectively utilize such software, especially when integrating it with existing educational systems. Besides, the long-term implications of integrating AI tools like MSAI in education include potential shifts in teaching methodologies and the need for ongoing teacher training to adapt to these technologies. Some teachers unfamiliar with Magic School AI may find it challenging to use this tool effectively. Unfortunately, a significant limitation of MSAI is its inability to offer detailed feedback and specific modifications for teachers’ lesson plans. While it provides general input and suggestions, the onus remains for teachers to manually revise their lesson plans. MSAI falls short in furnishing direct and actionable activities that teachers could readily implement in their curriculum.

7. Conclusions and Suggestions

The Magic School AI provides teachers with more support and resources in special education to modify their teaching and create an inclusive classroom. Especially a breakthrough function for teachers who are not teaching “highly-demanded subjects.” With the function of assignment scaffolders, text levelers, and text scaffolders, teachers can vastly reduce their workload by asking MSAI to modify the materials for them.

Moreover, Teachers usually do not receive professional support in special education, but some students with IEPs are in their classrooms. Teachers can use text scaffolders, sentence starters, exemplars, and non-exemplars as modifications and accommodations tools to include all students in the classroom and have equal access to the content. However, there are some potential improvements for Magic School AI:

• The Magic School AI may categorize the functions to be more user-friendly. The MSAI has embedded many functions, and teachers need to spend a lot of time locating the function(s) they want. It may be helpful if MSAI categorizes these functions as Special education, Math, ELA, parent communication, school reports, teaching tools, etc.
• It would be beneficial if Magic School AI could create tutorial videos to guide teachers with all the functions. This tutorial can serve as a mini professional development for teachers and help them understand how AI can ease their workload, create more accessible learning for all, etc.
• The Magic School AI could develop a more intellectual lesson plan modification system. The current functions generate broad modifications and accommodations. Teachers are still determining whether the lesson is inclusive enough for students with IEPs. If teachers can upload their lesson plans and receive direct feedback, modifications, and accommodations, they will be more confident in delivering their lessons effectively and efficiently.