Alexandra Webster-Birnberg

29 October 2024

The role of Artificial Intelligence (AI) in the classroom has sparked ongoing debate among educators. While public education in the US has functioned without artificial intelligence for the better part of the last three centuries, the question of its potential benefits in the general K-12 setting continues to spark discourse. Much of the discussion revolves around the potential for AI to create customized, interactive learning experiences, allowing for tailored education for each student. However,  these potential benefits are weighed against concerns about the risk of academic dishonesty– including cheating– facilitated by AI technologies that have stunted much potential progress. The public school system has seemingly yet to face an epic failure that AI could have solved. Or has it?

The Individuals with Disabilities Education Act (IDEA) was initially passed by Congress in 1975 under the name Education for All Handicapped Children Act and has a rich historical context. It has continued to be amended and expanded until its most recent update in 2004. This piece of legislation guarantees students with disabilities the right to “free and appropriate education” (FAPE), which is appropriate to the student’s educational needs and can be provided via a 504 or an Individual Education Plan (IEP). IDEA has its roots in Brown v. The Board of Education decision, where the Supreme Court struck down the doctrine of separate but equal as it applies to racial segregation in schools. Two subsequent cases in the early 1970s did their part in extending these protections to students with both mental and physical disabilities [1] .

At its core, IDEA strives to provide equal educational quality and opportunity to students of all means and abilities, protecting their right to be in the same classrooms as other students while still providing the necessary assistance for their individual cases. The passage of such secured necessary protections for students with different educational needs. In 2022, the number of students under IDEA qualified for an IEP, 504, or special education program sat around 7.1 million, or almost 15% of the overall student population. This is nearly double what it was in 1970, five years prior to the act’s passing [2] .

Although the existence of IDEA serves as a protection for the affected students on paper, a 2021 lawsuit shows that might not be the case. Individual plaintiffs, in cooperation with Disability Rights Texas, challenged the Austin Independent School District due to their failures to evaluate students who could be protected under IDEA in a timely manner due to backlogs of requests. In summary, by failing to provide students with the requested disability testing in a timely manner– as required by law–-they effectively restricted the speed at which students could qualify for IDEA protections. The resulting backlog left many affected students on long waiting lists for accommodations they rightfully deserved. The backlog revealed  systemic problem with the school district having accumulated hundreds of overdue evaluation requests [3]. 

Sadly, this is not a unique problem. The district’s vacancies cited as the root cause of testing delays extend beyond administrators of psychiatric exams. The Pew Research Organization highlights that whether a student is successfully recommended for special education resources, 504, or IEP directly correlates with a school’s socioeconomic makeup and resources. 

The report also notes that in the years 2020-2021, 40% of public schools not only reported having vacancies in their special education positions but also reported having significant difficulties filling said vacancies or did not successfully fill them at all [4] . What is the root cause of these vacancies? Lack of support specific to the needs of special education faculty. A new report by Frontline Education shows that, on average, a special education teacher can expect to spend less than half their days teaching. So, where is the rest of the time spent? It is largely spread across paperwork  [5]. On top of their ordinary workload, adapting complex lesson plans to the needs of individual students– which they have been trained to do– special education teachers must take particular care to maintain a record of their students’ progress. Tracking this progress can, in many cases, be the difference between finding a successful individualized learning method or not. Tracking progress means not just grading every assessment but also evaluating how each individual student’s results might progress them in their IEP goals, such as reassessing and making alterations when things don’t work and reinforcing the successes when they do. Even under ideal circumstances, individual teachers would have a hard time maintaining a deliberate understanding of each student’s educational abilities,. However, a systematic underfunding of special education programs coupled with rapidly growing class sizes has only underscored the severity and reach of the problem. Schools with the fewest resources are especially challenged  and are continuing to fail this already historically underserved population.

So, where does AI fit into all this? AI has the potential to significantly decrease the unmanageable workload of special education teachers while simultaneously providing better quality education via tailored support. This personalized approach of AI can make students feel understood and cared for, offering a promising future for special education regardless of school resources.

We’ve already seen examples of AI-based programs improving education in other contexts. Tools such as Grammarly and Turnitin are widely used for writing support and plagiarism detection, while platforms like Duolingo and -on our very own campus- ALEKS¹ provide customized learning experiences, tailoring lessons to individual progress. Extrapolating from these examples, AI can similarly enhance special education by automating time-consuming tasks like paperwork, giving teachers more time to work directly with students. Additionally, running sequences of student assignments through AI programs can better analyze patterns in student misunderstandings than any teacher could be expected to. Imagine a chronological portfolio of student writings, where the software identifies assignments and teaching methods that led to uncharacteristic gains in a particular area. The teacher could then use that information to adapt a student’s Individual Education Plan (IEP) to focus on strategies that are working. This level of personal analysis and attention to detail is not feasible without machine learning programs.

Additionally, AI can be used to adapt the same assessment to different skill levels. Take reading levels, for example. Currently, standard practice is providing different assignments to cater to different student abilities, which gives more work to the teachers who must create, provide instruction for, and grade  completely different assignments as well as austenitizing for students who are aware they are being given different material. With AI, a teacher could create one assessment, and the algorithm could adjust vocab and nuance accordingly. Such calibration would allow the instructor to make more efficient use of their limited time and resources.

For students with physical disabilities, like a visual impairment, AI Text to Speech (TTS) software can provide auditory assignment accommodation, rather than requiring a teacher to stop instruction to cater to a specific student’s needs. Human-like TTS software also provides autonomy to students with specific physical needs, giving them an experience more similar to that of their peers. In many cases, a program like this is the difference between a student being able to remain in a general education classroom or not. This is vital to student success; keeping students in the general education environment can have significant mental health and well-being benefits, fostering a sense of belonging and reducing feelings of isolation.

As legislators and school districts alike propose sweeping bans and stunting policies related to the use of AI, we must ask ourselves who the real stakeholders in these policies might be.  They’re not the kids in general classrooms but rather the individuals in special education environments who have the most to lose in terms of quality of education. As we move forward, policymakers will be increasingly asked to act on the impacts of newfound technology. Prior to making these decisions, we must look at the broader implications of these tools as apparatuses of equitability within the school system. Policies that prematurely prohibit  AI don’t just stomp progress but actively widen the gap for the students left behind.

As we move forward with discourse surrounding the use of generative artificial intelligence in the classroom, it is essential to look at its applications beyond the General Education Classroom setting and think about the people who could benefit most from this newfound resource. Additionally, we must identify that equity in education quality should be the foremost concern of public schools, both from the general classroom to a special education classroom setting, and from every special education classroom across the country as we seek to create a world where socioeconomic status, geographical location, and faculty resources do not determine student success; because When these students are provided the resources they need, success is within reach.

---

Image via Pexels Free Photos

Works Cited

[1] Disability Justice. 2023. “The Right to Education – Disability Justice.” June 1, 2023.  https://disabilityjustice.org/right-to-education/.

[2] Klein, Alyson. 2024. “The Pros and Cons of AI in Special Education.” Education Week, May 17, 2024. https://www.edweek.org/teaching-learning/the-pros-and-cons-of-ai-in-special-education/2024/05.

[3] Bone, Caroline, and Constance Smith. 2024. “Artificial Intelligence in Special Education | Frog, Part of Capgemini Invent.” Frog, Part of Capgemini Invent (blog). February 1, 2024. https://www.frog.co/designmind/artificial-intelligence-in-special-education.

[4] Acsa Profit. 2024. “Leveraging AI in Special Education.” Leadership Magazine. April 5, 2024. https://leadership.acsa.org/leveraging-ai-in-special-education.

[5] National Center for Education Statistics. n.d. “Enrollment in Public Elementary and Secondary Schools, by Region, State, and Jurisdiction: Selected Years, Fall 1990 Through Fall 2023.” https://nces.ed.gov/programs/digest/d13/tables/dt13_203.20.asp.

[6] Blazina, Carrie. 2024. “What Federal Education Data Shows About Students With Disabilities in the U.S.” Pew Research Center, April 14, 2024. https://www.pewresearch.org/short-reads/2023/07/24/what-federal-education-data-shows-about-students-with-disabilities-in-the-us/.

[7] Frontline Education. 2024. “Special Education Paperwork: How Much Time Does It Really Take?” June 6, 2024. https://www.frontlineeducation.com/solutions/special-ed-interventions/insights/reclaim-teacher-time/.

[8] Pendharkar, Eesha. 2023. “The Number of Students in Special Education Has Doubled in the Past 45 Years.” Education Week, August 1, 2023. https://www.edweek.org/teaching-learning/the-number-of-students-in-special-education-has-doubled-in-the-past-45-years/2023/07.

[9] “J.R. V. Austin Independent School District | Civil Rights Litigation Clearinghouse.” n.d. https://clearinghouse.net/case/18556/.

[10] “Research Behind ALEKS.” McGraw Hill aleks. Accessed October 17, 2024. https://www.aleks.com/?_s=6864838192166828. 

[11] Lee, Andrew M.I., JD. 2023. “IDEA, Section 504, and the ADA: Which Laws Do What.” Understood. December 4, 2023. https://www.understood.org/en/articles/at-a-glance-which-laws-do-what.

[12] “Districts Are Failing Special-needs Students. School Choice Is Helping.” n.d. The Thomas B. Fordham Institute. https://fordhaminstitute.org/national/commentary/districts-are-failing-special-needs-students-school-choice-helping.