AI For Innovative School-Based OT: Practical Tools For Enhanced Student Support

Editor's note: This text-based course is a transcript of the webinar, AI For Innovative School-Based OT: Practical Tools For Enhanced Student Support, presented by Julia Colman, OTD, OTR/L, BCP.

Learning Outcomes

• After this course, participants will be able to identify key AI tools that can support occupational therapy (OT) documentation, assessment, and intervention in school settings.
• After this course, participants will be able to recognize the benefits and challenges of integrating AI into school-based OT practice.
• After this course, participants will be able to list practical AI applications that enhance student support and streamline OT workflows.

Introduction

As an occupational therapist and an MBA student, I have become increasingly interested in how artificial intelligence (AI) can enhance efficiency in our field. The ability to streamline processes and improve the quality of services we provide is something I find both exciting and necessary. Whether we are working with students, patients, or clients in various settings, the potential for AI to support and optimize our work is undeniable. For this course, we will focus specifically on documentation and how AI can enhance accuracy and efficiency in this crucial aspect of our practice. Additionally, we will explore how AI can help us develop more personalized interventions within school-based settings.

This is a beginner-level course designed to introduce fundamental concepts of AI and its applications in occupational therapy. I recognize that there are varying levels of familiarity with artificial intelligence, and given how rapidly this technology is evolving, it is important to approach this topic with an open mind. AI is continuously advancing, and how it can support our work are constantly expanding. By the end of this course, I hope to provide a strong foundational understanding of AI’s role in occupational therapy documentation and intervention planning, ensuring that we are equipped to integrate these advancements into our practice effectively.

What Is Artificial Intelligence (AI)?

Artificial intelligence (AI) is the simulation of human intelligence in machines designed to perform tasks such as learning, reasoning, and problem-solving. While there are various definitions of AI, this is the working definition we are using for the purposes of this course. We are discussing systems that perform tasks typically requiring human intelligence.

What we know about artificial intelligence is that these systems have the ability to adapt and improve their performance over time. They do this through algorithms, data processing, and advanced computational techniques, all of which work together to simulate human cognitive functions.

Different Types of AI

In this course, I will define the different types of artificial intelligence and provide specific examples to illustrate each one. This discussion will dive deeper into the various forms of AI that currently exist, moving from simpler systems to more complex ones.

As we progress, we’ll examine these types in order of increasing sophistication. Starting on the left in Figure 1, in the gray box, we see narrow or rule-based AI. This type of AI operates within predefined parameters and follows specific rules to complete tasks. It is designed for particular functions and does not generalize beyond its programmed capabilities.

Figure 1. Different types of AI (Click here to enlarge the issue.).

We refer to systems that focus on specific tasks by following predefined rules when discussing narrow or rule-based AI. These systems are widely used, indicated by the two asterisks, signifying their long-standing presence in various applications. Because they rely on explicit programming, they cannot learn or adapt beyond their given parameters.

Moving into machine learning, we see a subset of AI where systems learn from data without being explicitly programmed. These systems can identify patterns, make predictions, and improve their performance over time, making them more advanced than narrow AI. Like rule-based AI, machine learning is also widely used and is not considered a novel technology compared to recent developments in artificial intelligence.

Deep learning takes machine learning a step further, utilizing neural networks to mimic human brain function. This makes it a more sophisticated and complex form of AI. Unlike rule-based AI and traditional machine learning, deep learning is not as widely used, though it is actively applied in fields like education and healthcare. As indicated in the slides, deep learning has only one asterisk, signifying that while it is in use, it is not yet as prevalent as the earlier forms of AI.

Natural language processing (NLP) is a specialized AI designed to understand and generate human language. As we discuss these different types of AI, I encourage you to think about some AI systems you are already familiar with or perhaps even use in your practice. These technologies fall into different categories, and in the next slides, I will provide examples that illustrate how they function in real-world applications.

Adaptive learning systems represent another advancement in AI, adjusting content delivery based on user performance and individual needs. Because these systems require more complex algorithms than NLP, they involve more technological sophistication. Despite their complexity, they are widely used, and I will provide well-known examples that you may already be familiar with.

Finally, we come to generative AI, which includes technologies like ChatGPT. This form of AI is designed to create new content, whether text, images, or code. Unlike rule-based AI, which follows predefined instructions, generative AI can perform higher-level cognitive functions that resemble human creativity and reasoning, producing entirely new and original material.

Importance of AI Literacy, Training, and Adoption for School-Based OTPs

Before diving into specific details and exploring how AI can be integrated into school-based occupational therapy, I want to take a moment to discuss why this conversation is so important. AI is everywhere—in the news, on social media, and across various industries. It is not just impacting education; within school-based occupational therapy, AI has the potential to be highly relevant and transformative.

One of the most significant reasons we need to discuss AI today is that it can enhance decision-making and efficiency for occupational therapy practitioners (OTPs). When we think about school-based caseloads, the reality is that they are often overwhelming. Many therapists are responsible for large numbers of students across multiple schools, making it difficult to prioritize and track each student’s needs effectively. AI can help streamline these processes, allowing us to organize our workload more efficiently and ultimately freeing up more time to focus on the quality of our services rather than getting lost in administrative tasks.

Another key advantage is AI’s ability to enhance the personalization of interventions. Many AI tools can analyze data to help tailor therapy plans to meet the specific needs of each student. We are all familiar with resources such as intervention session banks, goal banks, or quiz banks—broad tools that can be applied to multiple students. However, AI can help us refine these tools further, allowing us to adjust a general goal based on individualized data, ensuring that our interventions are more precise and effective.

Research also supports the positive impact AI has on diverse learners, particularly students with Individualized Education Programs (IEPs) and those receiving special education services. These students benefit from AI-enhanced interventions, which improve both the quality and accessibility of the services they receive. Additionally, AI can help automate administrative tasks, such as documentation, a time-consuming aspect of school-based practice. Progress tracking, in particular, is an area where AI can provide more objective insights. As therapists, we often ask ourselves whether the methods we use to track progress reflect meaningful change. AI can help answer that question by providing more consistent and data-driven measurements of student outcomes.

For many therapists, writing reports is another area where AI can be incredibly beneficial. Strong documentation is essential, but not every therapist feels confident in their writing abilities. AI can assist in crafting professional, well-structured reports and IEPs, ensuring that documentation is clear, comprehensive, and accessible to the rest of the educational team. This improves communication and ensures that the plans outlined in these reports are easily understood and followed.

One of the biggest benefits of AI in school-based OT is that it allows us to spend more time with students. By reducing the burden of administrative tasks, AI enables us to dedicate more time to direct student interactions and teacher consultations. Additionally, AI-driven predictive analytics and machine learning models can help us identify patterns in student performance, allowing for earlier interventions. While we have the skills to analyze data and recognize patterns ourselves, the sheer volume of students on our caseloads makes it nearly impossible to do this efficiently without the assistance of AI. These tools allow us to intervene sooner with students who are at risk and ensure they receive the necessary support.

AI also plays a significant role in improving accessibility for students with disabilities. It can be integrated into universal design principles to enhance communication and engagement for students with physical, sensory, or cognitive challenges. From speech-to-text applications to adaptive learning technologies, AI tools are expanding access and independence for students in ways that were previously not possible.

Beyond direct intervention, AI supports evidence-based practice by allowing us to analyze large datasets quickly. Instead of manually searching multiple databases, AI can efficiently gather relevant research, helping us stay current on best practices and innovative strategies tailored to specific student challenges. This ensures that our interventions are grounded in the latest research and aligned with high-quality, evidence-based approaches.

Creativity is another area where AI is proving to be a valuable tool. Instead of repeating the same activities in therapy sessions, AI can help generate fresh, innovative interventions. For example, if a student is working on fine motor skills and previously completed a lacing activity, AI can suggest variations of that activity to introduce a new challenge while maintaining engagement. This keeps therapy sessions dynamic and interesting for students.

AI also facilitates interdisciplinary collaboration by making it easier to share data and communicate with other team members, including educators and parents. Quick access to AI-generated home programs ensures they are well-written, individualized, and accessible for families, strengthening communication between therapists and caregivers.

Whether we choose to embrace AI personally or not, the education system is already moving in this direction. As OTPs, it is essential that we stay informed and up to date with these trends to remain relevant within the broader educational landscape. Schools are actively integrating AI into various aspects of education, and if we are not engaged in these conversations, we risk falling behind.

We must also address the ethical implications of AI use in education and therapy. This is not a simple topic—serious concerns are related to privacy, bias, equity, and accessibility. As we incorporate AI into our work, we must ensure that we use it in a fair and inclusive way for all students. We will discuss these ethical considerations further, focusing on implementing AI responsibly.

One of AI's most exciting opportunities is the potential to empower students toward greater independence. Promoting independence is central to what we do. AI can provide students with tools to support executive functioning, sensory needs, and adaptive skills without relying solely on one-on-one human assistance. Equipping students with AI-driven supports, we help them develop self-sufficiency and confidence in managing their needs.

Additionally, AI literacy is becoming essential in today’s evolving workplace. As educators and therapists, we have a role in advocating for equitable AI access for students, regardless of socioeconomic background. AI aligns well with the principles of universal design for learning by fostering engagement, representation, and expression, making education more inclusive and accessible.

Lastly, the increasing reliance on AI within education is also a response to financial constraints. Supporting students with diverse needs is costly, and school districts are looking for cost-effective ways to enhance services. AI is becoming a practical solution to help bridge these gaps. The rapid advancement of AI means that we, as practitioners, must remain adaptable and informed to ensure our practice remains relevant and impactful in this changing landscape. The world is evolving, and we must evolve with it as occupational therapy practitioners.

Practical AI Tools for OTPs

Now we’re getting into the part that most of you are likely here for—the practical AI tools that OTPs can use in their practice. As we look at these slides, I’ll be going through different types of AI and how they can be applied in OT settings. There are four slides covering various AI tools, and while I won’t be reading every single detail, I wanted to provide a comprehensive overview with plenty of examples. This is not an exhaustive list, as countless AI tools are available, but I will highlight some of the key ones, including those I have used in practice.

Each column in these slides presents specific information. I will introduce the type of AI, followed by examples of platforms or resources currently being used. I will also briefly touch on how each AI tool is relevant to occupational therapy and what client factors or performance skills these technologies support.

Starting with the first category, rule-based or narrow AI, these systems function within a predefined set of rules to complete specific tasks. These tools do not adapt beyond their programming but can be incredibly useful for automating repetitive tasks, enhancing efficiency, and supporting clinical decision-making. I'll discuss some of the most relevant examples and how they can be integrated into school-based OT and other practice areas.

As we move through these AI tools, remember that we are progressing from simpler technologies to more complex ones. Regarding rule-based AI, we refer to basic chatbots, embedded calculators, early diagnostic tools, scheduling and classroom management tools, basic spelling and grammar checkers, flashcards, and educational apps. These technologies follow predefined rules to complete tasks without learning or adapting beyond their programming.

Some tools in this category may already be familiar, such as Quizlet and ABC Mouse. If you work in special education, you’ve likely encountered tools like Google Read & Write or Natural Reader. Classroom behavior management tools, like Class Dojo, also fall into this category. What’s important to note is that many of these technologies were not originally AI-based when they were first developed. However, as AI capabilities have expanded, many existing tools have integrated AI-driven features, transforming how they function. For example, if you were familiar with Google Read & Write a few years ago, it’s worth considering how the tool has evolved and what AI-powered enhancements it now offers. Staying informed about these advancements is part of our ethical responsibility.

These rule-based AI tools can significantly enhance efficiency in screening and evaluation processes. Tools like Grammarly can streamline documentation, while classroom behavior management platforms such as Class Dojo provide data useful for tracking student progress and improving communication between therapists, teachers, and parents. Rather than reinventing the wheel, we can explore ways to integrate the data from these platforms into our workflow to enhance collaboration and efficiency.

Incorporating tools like Quizlet or ABC Mouse in intervention sessions can create more interactive and engaging student experiences. These tools can also support caseload management—Grammarly helps improve written communication, while visual timers and calendar reminders can assist in structuring sessions and keeping track of schedules.

Considering how these tools align with OT practice, consider executive functioning and cognitive skills. Visual timers, for example, are widely used in classrooms to support students in managing transitions and staying on task. Many of you are likely already using them in therapy sessions. Even if you have never considered yourself someone who actively integrates AI into your practice, if you use visual timers, you are already leveraging AI-driven technology to support student success.

Machine learning is another category of AI that has significant potential in school-based occupational therapy. Some commonly used examples include i-Ready, Reflex Math, and PowerSchool Analytics. I want to highlight PowerSchool Analytics, even though I have not personally used it. This is a school performance monitoring tool that many districts are beginning to adopt. Their website provides in-depth information about their software, and I encourage you to explore it.

PowerSchool Analytics uses predictive analytics to assist with IEP development, leveraging data-driven insights to create individualized plans and inform intervention strategies. This allows us to ensure that interventions align with a student’s goals and contribute meaningfully to their success in school. By utilizing platforms like this, occupational therapy practitioners can advocate for more data-driven IEP development, reinforcing the importance of tailoring interventions based on real student performance data. These tools can support executive functioning, cognition, visual processing, sensory processing, behavior, social participation, and more—making them valuable across many areas of student development.

Deep learning tools take this a step further by using advanced AI techniques like neural networks to recognize patterns in data. Some of the most relevant applications in special education include image recognition systems and accessibility tools that improve engagement and participation for students with disabilities.

One example is Otter.ai, which transcribes spoken language into real-time text. This is incredibly beneficial for students with auditory processing disorders, hearing impairments, or learning disabilities, as it provides an instant visual representation of verbal instructions or lectures. If a student struggles to process spoken instructions and frequently experiences behavioral challenges due to misunderstandings, having a text-based version of those instructions can reduce frustration, improve compliance, and decrease reprimands throughout their school day. This accessibility tool aligns perfectly with occupational therapy’s role in supporting executive functioning, self-regulation, and classroom participation.

Another promising tool is Seeing AI, which functions oppositely—it converts visual information into auditory descriptions. Designed for students with visual impairments, this app allows users to take a picture of an object or scene, and it will provide an auditory description of what was scanned. While working in the school system, I saw firsthand how challenging it was for therapists, teachers, and teams to create effective accommodations for students with low vision. The accessibility gap was a real struggle. Tools like Seeing AI to support these students in real time make classrooms more inclusive, accessible, and adaptable to diverse needs.

Another AI-powered tool, ModMath, integrates handwriting recognition to convert handwritten math problems into digital text. Many students with visual processing differences struggle with reading certain fonts or deciphering handwriting, which can make math assignments unnecessarily difficult. By digitizing handwritten work, this tool enables students with fine motor, visual-motor, and visual processing challenges to complete math assignments in a format that works for them. While OTPs primarily work with students receiving special education services, these technologies can benefit all learners in the classroom.

Beyond individual interventions, AI tools like these help advance universal design for learning. As occupational therapy practitioners, we play a crucial role in advocating for and guiding the integration of these technologies into classrooms. By working alongside teachers and school staff, we can promote inclusive educational strategies that meet the needs of more learners while fostering accessibility and engagement for all students.

Natural language processing (NLP) encompasses a wide range of AI tools, and while I won’t go through every single one, I want to highlight a few key examples that are particularly relevant to occupational therapy practice. Virtual tutors and certain types of communication devices for non-speaking students fall under this category, as well as speech-to-text tools, which many of you are already using. If you have incorporated speech-to-text technology into your practice, you already engage with AI.

Virtual assistants like Siri, Alexa, and Google Assistant are excellent examples of affordable and readily available NLP-based AI tools. I often encourage their use, especially in adaptive classrooms focused on life skills development. These tools can support students in completing daily tasks, helping to enhance their independence. Bringing virtual assistants into the classroom can be incredibly beneficial, providing consistent, structured support that aligns with the needs of students with cognitive or executive functioning challenges.

Another application is social skills training tools. AI-powered chatbots like Replika are designed to simulate human conversation, making them particularly useful for students who struggle with navigating social interactions. OTPs typically use social stories or role-playing exercises to help students develop these skills. However, AI can provide a much broader range of dialogue and scenarios than we, as therapists, might generate. When we create social scenarios, they are often shaped by our own experiences and biases. AI, while not free from bias, provides a more diverse range of interactions that may offer new learning opportunities for students. This makes AI a valuable tool in supplementing traditional social communication interventions.

Language translation tools such as Google Translate and Microsoft Translator support English as a second language (ESL) students. These AI-driven tools help bridge language barriers and enhance learning for students requiring additional language support.

For those looking for specific speech-to-text tools, some of the most widely used include Dragon NaturallySpeaking, Kidsense AI, and Google Speech-to-Text. These tools convert spoken language into written text, making them beneficial for students with writing challenges, executive functioning difficulties, or other language-based disabilities. Similar to other AI applications, NLP tools enhance accessibility, can be incorporated into direct OT sessions, and can also be recommended as part of home programs.

One of the most impactful ways NLP-based AI can support school-based occupational therapy is by helping with goal writing, documentation, and IEP generation. For students who benefit from speech-to-text technology, it is essential to ensure that these accommodations are included in their IEPs. The ability to dictate assignments, generate written work, or use voice commands can be a game-changer for students with written expression difficulties.

These AI tools have a broad range of applications within school-based practice. Whether they improve student accessibility, facilitate communication, support social interactions, or enhance documentation, NLP-based AI is becoming integral to the educational environment. As OTPs, recognizing and integrating these tools into our practice ensures that we leverage available technology to provide the best possible support for our students.

Adaptive learning systems are designed to personalize instruction based on a student’s needs. Some examples of these tools include Dreambox, Lexia, and Amplify Reading. These platforms analyze student performance data and adjust the difficulty of instruction accordingly, either increasing or decreasing the level of challenge to match the learner’s abilities. This ability to modify instruction in real-time makes these tools valuable for students who need more tailored support, particularly in areas like reading or math.

One area where virtual reality is proving particularly useful is in the education of occupational therapy students. Virtual reality can also be a powerful tool for professional development within school districts. For example, if a district focuses on social-emotional learning, VR could create case-based learning experiences for teachers. It could also provide an interactive way to educate staff about sensory processing challenges in the context of social-emotional learning.

Computer-assisted cognitive training is another category of adaptive learning technology that can be integrated into therapy. Tools like CogMed and Lumosity offer cognitive skill training through app-based exercises. These can be incorporated directly into therapy sessions or assigned as home programs to provide students additional practice in memory, attention, and problem-solving.

Eye-tracking technology, such as Tobii, is another adaptive learning tool for augmentative and alternative communication (AAC). This allows non-speaking students with significant motor impairments to communicate using eye movements, giving them greater independence in their interactions.

Adapted games are also key to making learning accessible and engaging for diverse learners. As a neurodiversity-affirming therapist, I prioritize strength-based therapy incorporating a student’s interests. Using adapted games designed for accessibility allows students to engage more fully in therapy sessions and the classroom environment. Minecraft: Education Edition is one example of an educational tool that has been enhanced with accessibility features, making it more inclusive for neurodivergent students and those with different learning needs.

Generative AI is the final category I want to discuss, as it is rapidly expanding its role in education and occupational therapy. Some well-known examples include ChatGPT for text generation and DALL·E for image generation, developed by OpenAI. AI-powered content creation tools can significantly enhance caseload management, productivity, and intervention planning.

Canva’s AI design tool is an example of how AI can reduce the time spent developing presentations, handouts, and educational materials. One of the major challenges in school-based OT is ensuring that teachers and other team members are informed about the interventions being used in therapy. Using tools like Canva’s AI-powered features makes creating visually appealing and informative resources easier, helping maintain communication and collaboration across the educational team. The benefit is that these tools allow for customization and personalization while significantly reducing the time and effort required to create materials.

Other AI-driven resources, such as Kahoot’s AI quiz features and Speechify for text-to-speech conversion, further support student engagement and therapist efficiency. AI tools like these enhance accessibility, streamline documentation, and improve intervention planning, freeing more time for direct student interaction and individualized therapy.

Case Application 1: Using AI to Personalize MTSS Interventions

Now, we are going to switch to a case application. This one examines how to use AI to personalize multi-tiered systems of support interventions, as shown in Figure 2.

Figure 2. Multi-tiered systems of support interventions.

I will go through all the different levels and discuss how we can use these things across all the different tiers within the school system.

In this first example, the OTP is working in a public middle school where the student population has diverse needs, including sensory processing challenges, learning disabilities, and emotional regulation difficulties. The OTP collaborates with teachers, counselors, and families to support students' participation in academic and non-academic activities.

At the population level, looking at that first tier of universal interventions, the school has noticed an increase in students experiencing stress and dysregulation during transitions, such as moving between classes. These challenges are affecting overall engagement and learning. If you work in the school system, this situation is real and something we are seeing increasingly. Students are struggling with emotional dysregulation and difficulty with transitions.

Integrating AI at this universal level, the OTP could use an AI-powered survey tool to gather data from teachers about daily stressors and transition challenges. Data can also be collected from students, allowing for the identification of common barriers. The AI could then analyze these responses, revealing patterns such as noise levels in hallways or insufficient transition time. Based on this data, the OTP could develop personalized solutions, such as playing calming music in hallways or displaying a visual timer on hallway screens to provide clear expectations for students. This promotes a sensory-friendly transition protocol at the universal level, ensuring that all students benefit from a structured and supportive environment.

At the targeted level, this scenario focuses on a group of sixth graders who have experienced bullying and are struggling with emotional regulation and executive functioning. Research shows a link between bullying and academic challenges, particularly in areas related to executive function and self-regulation. In this case, the students struggle with writing tasks and completing assignments.

To integrate AI in this scenario, the OTP recommends that the district provide access and training to these at-risk students on using an AI-powered writing app, such as the Google Read and Write extension. This tool offers personalized supports like sentence starters, vocabulary suggestions, and text-to-speech features, reducing the stress associated with writing tasks.

Additionally, pairing the writing app with an emotion-tracking tool allows students to log their feelings before and after completing specific tasks, particularly non-preferred ones. AI could then provide real-time coping strategies such as guided breathing or movement breaks based on the student’s emotional state. Unlike traditional low-tech strategies, where students must independently select a coping mechanism, AI can generate tailored responses that support emotional regulation in the moment.

The anticipated outcome of these interventions is improved work completion and increased confidence in writing tasks, leading to greater academic success and emotional well-being.

At the individual level, the case involves a seventh-grade student with ADHD and sensory processing challenges who struggles to stay engaged in class due to difficulties regulating attention and sensory input.

One AI-based intervention for this student could be using an AI-powered wearable device. Many fitness trackers already monitor physiological responses like heart rate and skin temperature, providing early sensory overload or inattention indicators. Given how often we discuss dysregulation and interoception in OT, this technology could be a valuable tool to support self-awareness of physiological responses.

The OTP could also integrate AI-driven timers to prompt sensory breaks when needed. AI could further help customize the student's daily schedule by analyzing physiological data to determine when they are most dysregulated. Instead of providing movement breaks at arbitrary times, this approach ensures that sensory breaks are scheduled when they will be most beneficial.

As this example illustrates, AI can be used across all levels of intervention to support student learning. From universal supports that benefit the entire school community to individualized interventions tailored to specific student needs, AI offers a powerful tool for improving engagement, accessibility, and self-regulation in school-based occupational therapy.

Case Application 2: Using AI to Support Caseload Management

The next scenario is a case application of AI to support caseload management, focusing on how AI can help therapists streamline their workload.

The occupational therapist manages a diverse caseload of preschool, elementary, and high school students across four schools in this scenario. If you are a school-based therapist, this likely sounds very familiar. The therapist’s responsibilities include direct individual push-in services, group pull-out services, and consultative support. Balancing these demands requires efficient scheduling, data tracking, and resource development.

AI can assist with this through the Microsoft Outlook Calendar with its AI assistant. This tool provides automated scheduling that optimizes the therapist’s time. By inputting service requirements and availability, the AI assistant can generate an efficient schedule that minimizes travel time, ensures a balance between direct, group, and consultative services, and automatically suggests optimal time slots. Anyone who has had to create a school-based therapy schedule knows how time-consuming and complex this process can be. It often feels like playing a game of Tetris, trying to fit everything together in the best way possible. AI can help reduce the time spent on this task while improving accuracy and efficiency. The AI assistant can also send alerts for overlapping commitments and provide updates on travel delays, ensuring that therapists remain organized and on track throughout their day.

Another way AI can support caseload management is through platforms such as TheraPlatform and Goalie, specifically designed for therapists to document session notes more efficiently. In healthcare settings, AI-driven tools for documentation, such as voice dictation and automatic SOAP note generation, have already been widely implemented. Similar AI-supported tools are now available for school-based therapists, making it easier to complete documentation and track progress over time. AI-generated data can help therapists visualize student progress, making it easier to identify patterns and make data-driven decisions about intervention strategies.

AI can also be a valuable tool for content creation. It can assist in developing customized sensory plans, social-emotional learning activities, and fine motor exercises that can be completed outside direct therapy sessions. Generative AI can generate innovative and engaging ideas, while design tools like Canva’s AI-powered features can create visually appealing and user-friendly handouts for teachers, parents, and students.

Visual schedules are another example of how AI can enhance caseload management. AI-generated templates can create personalized schedules for students, helping them understand their daily routines and anticipate their OT sessions. A simple handout that outlines when the student has OT each week can foster greater independence, allowing them to prepare for sessions without relying entirely on adult prompts. AI-generated resources like these support therapists and students by improving organization and engagement in therapy.

Case Application 3: Using AI to Write Strengths-Based Goals

This is our last case, focusing on using AI to write strengths-based goals.

In this case, the IEP team is meeting to draft an annual IEP for a fifth-grade student. The student has strengths in verbal communication, visual memory, and creativity, while the areas of need include writing mechanics, organizing thoughts into paragraphs, and sustaining attention during writing tasks.

The special education teacher wants to include a writing goal for the student to compose a five-sentence paragraph independently by the next IEP. The occupational therapist supports the student with executive functioning, fine motor coordination, and self-monitoring strategies.

Using ChatGPT, we can develop a strengths-based IEP goal or receive support in refining one. AI can generate goal suggestions that incorporate the student’s existing strengths while addressing areas of need in a way that promotes engagement and motivation. This allows for a more personalized and student-centered approach to goal writing.

I provided an example of a prompt that could be entered into ChatGPT to help generate a well-structured, strengths-based IEP goal.

Figure 3. Example ChatGPT prompt.

So I said, help us write a strengths-based IEP writing goal for a fifth-grade student.

You never want to input specific student data into ChatGPT or any AI tool. That means avoiding personal identifiers like the student's name, birth date, or other sensitive information. However, you can include more generic details without compromising student privacy.

For this case, I provided the following prompt: The goal should focus on composing a five-sentence paragraph by the end of the year. The baseline data indicates that the student can copy up to three sentences with a visual model and adapted paper. The goal should incorporate the student’s strengths in verbal communication, creativity, and visual memory while addressing areas of need, such as organizing ideas, fine motor skills, and self-monitoring. It should also include measurable criteria and emphasize leveraging the student's strengths.

This was the generated goal:

By [date], the student will independently compose a five-sentence paragraph that includes a topic sentence, three supporting details, and a conclusion, using a graphic organizers, self-monitoring checklists, and keyboarding tools as needed. The paragraph will demonstrate 80% accuracy in organization, punctuation, capitalization, and spacing in 3 out of 4 trials, as measured by teacher observation and work samples.

As you can see, this goal effectively leverages the student's strengths while incorporating the necessary supports. It provides clear, measurable criteria, making it easier for the OT to refine specific intervention strategies that align with the student’s needs.

Given this information, the IEP team can confidently draft an appropriate goal for the student. If they develop shared goals, the OT can collaborate with the special education teacher. This allows the OT to support the student’s creativity and visual memory while ensuring that the writing goals set by the special education teacher are effectively addressed.

Key Advantages

Looking at the evidence, the key advantages of using AI are again centered around efficiency. AI reduces the time spent on documentation, allowing for more direct student interaction. It also enables more personalized interventions by adapting recommendations based on individual student data.

I have included the references here because I want you to see that there is evidence supporting these claims. This is not something I made up—there is research backing the integration of AI in education and therapy. The need to incorporate these tools is clear based on the data available.

Another key advantage is enhanced accuracy. AI minimizes human error in documentation, a well-documented benefit across healthcare and educational settings. Improved accuracy ensures that reports, progress notes, and IEP goals are clearly written, consistently formatted, and reflect student progress. This level of precision contributes to more effective intervention planning and collaboration among team members.

Challenges and Barriers

But looking at some challenges or barriers, one major concern is protecting sensitive student information. As mentioned, you should never input specific student identifiers into AI tools like ChatGPT. Ensuring that student data remains confidential is essential, and school-based occupational therapists must be cautious about how they engage with AI to maintain compliance with privacy regulations.

AI also contains bias. The data and algorithms that drive AI tools can reflect existing biases, which can influence recommendations and outcomes. Addressing this risk requires critically evaluating AI-generated outputs to ensure they align with ethical and equitable practices.

Another challenge is the training required to use AI tools effectively. AI is complex, and as occupational therapy practitioners, we are not data analysts or IT specialists. It is essential to ensure that OTPs have adequate training and support to integrate AI meaningfully into their practice. Without proper understanding, AI tools may not be used to their full potential or implemented in less effective ways.

Cost is another consideration. Many AI-powered platforms come with a price, and some can be expensive. Schools and districts must determine how to allocate resources effectively and select the most appropriate software programs. Balancing cost with benefit is an ongoing challenge, particularly when advocating for adopting new technology in educational settings.

AI & AOTA Code of Ethics

I wanted to bring in the AOTA Code of Ethics here. You guys are very familiar with this. I'm not going to read it verbatim, but you can see that across the different ethical responsibilities or ethical principles that we have as therapists, there are ethical implications when it comes to AI.

When we're talking about beneficence, or acting in the best interest of students, using AI is a way to ensure that we are doing just that. AI allows us to provide evidence-based interventions that positively contribute to students' well-being and success. By integrating AI tools thoughtfully, we can enhance the effectiveness and efficiency of our interventions while ensuring that students receive high-quality support tailored to their needs.

Non-maleficence involves taking action to prevent harm. This means being cautious in selecting AI tools, ensuring they are validated and have strong evidence supporting their use for the specific population of students we serve. Using AI tools lacking reliability or research backing can lead to inaccurate assessments and interventions, ultimately hindering student progress. We are responsible for critically evaluating these technologies and ensuring they align with best practices before integrating them into our work.

Autonomy is about respecting individuals' rights to make informed decisions regarding their participation. When using AI tools to collect and analyze student data, it is essential to obtain informed consent. This is particularly important for platforms like TheraPlatform, where data is stored and analyzed. Transparency with families and students about how their information is used ensures that they know the process and can make informed decisions about participation.

Justice relates to fair and equitable access, ensuring that AI is used to benefit all students who need it. Equity plays a crucial role here, as it is important to implement AI tools fairly, ensuring that every student who could benefit has access. Advocating for policies that minimize disparities in access to these technologies is essential in promoting justice. Integrating AI into school-based occupational therapy increases the potential to reach more students, making processes more efficient and ensuring that interventions are accessible to those who need them most.

Veracity is truthfulness, which means being transparent about the limitations of AI and avoiding any exaggeration of its accuracy. It is important to acknowledge that these are evolving technologies and to be truthful in how we interpret the output AI provides. This ensures that we maintain realistic expectations and do not over-rely on AI without critically assessing its recommendations.

Fidelity involves loyalty, respect, and commitment in our professional relationships. This means maintaining trust with colleagues, families, and students by ensuring that AI tools are used appropriately and within our scope of practice. It is essential to implement these technologies in ways that align with our professional training and the intended purpose of the tools. Using AI responsibly means not overstepping professional boundaries or applying these tools in ways beyond their validated or ethical use.

Dos and Don'ts

• DO:
  • Understand the purpose of the AI tool.
  • Be mindful of data security & privacy.
  • Ensure ongoing training & adaptation.
  • Be aware of bias and accuracy.
  • Use tools to increase accessibility.
  • Use as a bridge to improve multidisciplinary team collaboration.

Some key to-dos and what to avoid when implementing AI tools in occupational therapy practice.

Understanding the AI tool's purpose is important, ensuring that it aligns with intervention goals and supports student success. Be mindful of data security and privacy, ensuring AI tools comply with regulations and do not compromise student information. Ongoing training and adaptation are crucial, as AI constantly evolves, requiring therapists to stay informed on best practices. Awareness of bias and accuracy is essential—AI is not infallible, and critical evaluation of its output is necessary. AI should be used to increase accessibility, helping students engage with learning and therapeutic activities in meaningful ways. Additionally, AI can serve as a bridge for multidisciplinary collaboration, improving communication and coordination among team members.

What should be avoided includes using platforms that are not compliant with privacy and security standards. Confidentiality must be maintained, never enter personal student information into AI tools. Do not rely on AI-generated output as 100% accurate—professional reasoning is always required to interpret and apply the information appropriately. Do not assume bias does not exist in AI; it is present in all data-driven systems and must be recognized to avoid reinforcing disparities. Finally, do not recommend AI tools or their use without an ongoing plan for training and support for students and the staff who will be implementing the tool. Ensuring that AI is used effectively requires continuous education and structured implementation.

Take Action

1. Evaluate Needs: Identify areas where AI can alleviate workload or enhance outcomes.
2. Collaborate: Work with educators, administrators, and families to ensure alignment and transparency.
3. Select Tools: Choose AI platforms that align with school policies and ethical standards. Make sure you’re asking the “big questions” when making purchases.
4. Pilot Programs: Test tools on a small scale before full implementation.
5. Stay up to date with AOTA guidance related to AI use.
6. Implement strict data governance policies.
7. Engage in continuous training and collaboration with tech developers; Offer training for school staff to understand and support AI applications.

Some ways to take action include evaluating the needs in your specific setting to determine where AI could alleviate workload and enhance outcomes. Collaborating with educators, administrators, and families ensures alignment and transparency in implementing AI tools.

Selecting AI platforms that align with school policies and ethical standards is essential. Asking critical questions before purchasing ensures that the tools chosen are intentional and appropriate for the student population. Pilot programs can be an effective strategy—testing AI tools on a small scale before district-wide adoption allows for adjustments and informed decision-making.

Staying up to date with AOTA guidance on AI use is important, as recommendations and best practices are continually evolving. Implementing strict data governance policies protects student privacy and ensures compliance with legal and ethical standards. Engaging in continuous training and collaboration with tech developers allows for a better understanding of AI applications, and offering training to school staff helps them effectively support the AI tools being integrated into practice.

Future of AI in School-Based OT

As I have said, this is constantly changing. The future is still unknown, but we know that AI will continue to advance, and we need to stay engaged with its development.

In terms of what we can anticipate, we will likely see more tracking and documentation processes requiring minimal manual input, shifting towards AI-based documentation systems. AI will also continue to support universal design for learning and inclusive education, helping to create more accessible and individualized learning experiences.

Another potential development is general AI, or strong AI, which is not currently in use. At this point, it remains hypothetical or theoretical, but it refers to AI that can think, reason, and perform intellectual tasks like a human would. Ongoing research and ethical debate about what this could look like and how it might impact various fields, including education.

As a thoughtful question, what implications would general AI have for education? If we reach a point where AI can perform human cognitive tasks, how might that reshape the way we teach and support students? There are no definite answers, but it is worth reflecting on as AI evolves.

Summary

Exam Review

1) What does AI stand for?

A. Artificial Imagination

B. Automated Intelligence

C. Artificial Intelligence   Correct Answer

D. Advanced Interaction

2) What is a key benefit of integrating AI into school-based OT practices?

A. It replaces the need for OTPs.

B. It enhances OTP decision-making and efficiency.   Correct Answer

C. It guarantees perfect outcomes for every intervention.

D. It eliminates all documentation requirements.

3) What is a key advantage of AI?

A. Efficiency

B. Personalization

C. Enhanced accuracy

D. All of the above   Correct Answer

4) Which of the following is a challenge when incorporating AI into school-based OT?

A. Reduced time for student interactions

B. Ethical concerns regarding data privacy   Correct Answer

C. Increased paperwork for OTPs

D. Lack of AI tools available

5) When "taking action" with AI, the OTP should...

A. Identify areas where AI can alleviate workload or enhance outcomes

B. Work with educators, administrators, and families to ensure alignment and transparency

C. Choose AI platforms that align with school policies and ethical standards.

D. All of the above   Correct Answer

Questions and Answers

Are these AI tools free, or what are the general costs associated with using them?
This is a great question, and the answer is—it depends. Some AI tools are free, while more complex tools that provide data-driven insights, such as TheraPlatform and other documentation platforms, often have higher costs. Pricing can vary based on the number of users or students using the platform. For example, costs may differ if an entire district adopts the software versus an individual subscription. Some tools, like calculators or the Google Read and Write extension, offer free versions for educators and students, but they may require verification of educational status. Ultimately, the cost depends on the specific technology being used.

How can machine learning be used for intervention planning?
A: Machine learning can help with intervention planning by analyzing data collected from platforms like TheraPlatform. If the platform includes built-in insights, those can be directly used to guide intervention strategies. Alternatively, data from these platforms can be integrated into generative AI tools like ChatGPT to develop effective and innovative intervention plans. Therapists can create individualized, data-driven therapy strategies by combining machine learning with generative AI.

What are some specific wearable devices that are useful in therapy?
There are many wearable devices available, and their effectiveness varies. More advanced and accurate devices typically come with a higher cost. The Apple Watch is particularly useful because it integrates easily with a phone or computer to collect and analyze physiological data. This can help track heart rate, movement, and other metrics that may be relevant for self-regulation or executive functioning. However, ethical considerations around cost and accessibility must be considered when recommending these devices. As technology evolves, new options continue to emerge.

How far are we from robots in therapy?
That’s a big question, and honestly, I don’t know the exact answer—but many people are thinking about it. AI-driven robotics in therapy is something that researchers are actively working on. The future of AI in school-based occupational therapy is evolving, and while we’re not there yet, advancements in AI suggest that robotic-assisted therapy may become a reality at some point.

References

Please see the additional handout.