An Algorithm Can't Know a Child: AI and the Future of the IEP

Authors: Dr. Teri Marx and Dr. David Bateman

The field of special education was fundamentally reshaped when Congress enacted the Education for All Handicapped Children’s Act in 1975, now known as the Individuals with Disabilities Education Act, or IDEA. Before its passage, many individuals with disabilities were denied meaningful access to public education. Some were excluded from school entirely; many others were placed in programs that provided little meaningful instruction. IDEA fundamentally changed this reality.  

Rooted in the broader civil rights movement of the 1960s and 1970s, IDEA reflects the principle that disability should never be a basis for exclusion from educational opportunity. Today, the law protects nearly eight million children and youth with disabilities, ensuring they have access to a free appropriate public education (FAPE)—an education specially designed to meet their unique needs and prepare them for further education, employment, and independent living. 

IDEA requires schools to identify students who need specially designed instruction, develop an individualized education program (IEP), establish a team to make educational decisions, and ensure the student is making progress toward their IEP goals and in the general education curriculum. When these requirements aren’t met, parents can exercise due process rights. 

Confounding these realities is the chronic underfunding of IDEA at the federal level. Congress has never fulfilled its promise to fund 40% of the excess costs associated with special education, leaving state and local education agencies (SEAs/LEAs) to shoulder much of the financial burden. These entities must navigate chronic underfunding while also addressing persistent staffing shortages, the growing number of students requiring an IEP, surges in costly dispute resolutions, and the costs of educating students with disabilities. Just as SEAs and LEAs are being asked to do more with less, teachers are, too.  

AI and the IEP Process 

It is within this context—one marked by staffing shortages, growing workloads, and persistent financial pressures—that artificial intelligence (AI) has emerged as an increasingly attractive tool. According to a Gallup poll, the rate of K-12 education personnel using AI has more than doubled between 2023 and 2026, rising from approximately 20% to 56%. In special education, a survey by the Center for Democracy and Technology found 57% of teachers used AI to help them develop IEPs during the 2024-25 school year—an 18% increase from the previous year. 

Perhaps most strikingly, 15% reported using AI to generate an entire IEP from beginning to end. This use of AI should not be surprising. Special education personnel operate in environments characterized by increasing paperwork demands, staffing shortages, and persistent compliance concerns. For many educators, AI is viewed as a means of reclaiming time and increasing efficiency. 

The Legal Stakes of an AI-Generated IEP 

AI's growing presence in education may not be surprising. More concerning is the limited attention being paid to its potential legal implications, particularly when AI is used to develop IEPs that IDEA requires to be individualized to meet each student’s unique needs. As AI becomes more deeply embedded in the IEP process, districts must consider not only what the technology can do, but also how its use may affect their obligations under IDEA. 

An IEP is the primary mechanism for ensuring eligible students receive a FAPE, and disputes between schools and parents often center on whether that obligation has been met. Under IDEA, the FAPE standard includes procedural and substantive requirements. Procedural requirements include adhering to timelines for determining eligibility and developing an IEP; meaningfully involving parents in educational decisions; ensuring the participation of required IEP team members; including all required components in the IEP; and implementing the services and supports specified in the IEP as written. 

Substantive requirements, by contrast, focus on whether an IEP is reasonably designed to enable the student to make appropriate educational progress. Ways in which the substantive requirements may not be met include failing to address all areas of need within the IEP, developing IEP goals that are not challenging or ambitious, not adequately monitoring a student’s progress, or not adjusting the IEP if a student isn’t making progress.  

AI Doesn't Bear the Responsibility—Schools Do 

We already know that many teachers are turning to AI to help them develop aspects of the IEP, and many school districts across the country have adopted AI-driven platforms. In many cases, that adoption has outpaced policy or guidance—especially regarding the IEP. Even when policies are in place, they may not adequately address the potential risks that AI use can pose to the procedural or substantive elements required to ensure FAPE. This doesn’t place AI at risk of being sued. It places school districts at risk. The question isn’t whether AI can or should be used; it already is. The more important question is who is responsible when AI contributes to a poorly designed or legally deficient IEP? The answer is clear: not AI.  

Three Guardrails for Using AI Responsibly 

So, what guardrails can districts put in place to protect students’ rights, support educators, and reduce legal risks? 

1. Establish policies related to AI use. Develop clear guidelines to protect student privacy, check for bias, ensure personnel use only approved platforms, and review any AI-generated content for accuracy and legal compliance. And policies need to be enforced. Consider what might happen if, in the absence of policy, a teacher uses AI to develop an IEP and the family pushes for a due process hearing because their child’s goals aren’t aligned with their needs or aren’t monitored appropriately. Who bears responsibility? Develop policies that plan around worst-case scenarios.  
2. Invest in ongoing professional learning. AI may streamline certain tasks but cannot substitute professional judgment—nor can it actually implement an IEP once it’s developed. Ensuring administrators and educators are trained to develop and implement a high-quality, legally sound IEP serves as a protective factor. The second someone can’t answer why something is in the IEP, how to implement it, or how to monitor a student’s progress toward their goals, safeguards intended to protect students—and schools—begin to break down.  
3. Advocate for IDEA to be fully funded. Part of AI’s appeal is practical: districts are being asked to serve increasing numbers of students with disabilities while managing persistent staffing shortages, heavier workloads, and limited resources. Educators are expected to meet increasingly complex instructional, procedural, and documentation demands with too little time and capacity. Fully funding IDEA would not eliminate the need for thoughtful use of technology, but it would help address some of the underlying conditions that make technological shortcuts so attractive. AI should supplement professional capacity, not replace the personnel and resources needed to provide students with the services IDEA requires.   

Technology Cannot Replace Human Judgment 

Ultimately, the student must remain at the center of every decision. IDEA’s promise is fundamentally human: educational decisions affecting children with disabilities should be individualized, collaborative, and informed by people who know the student, understand the evidence, and are accountable for the results. An algorithm cannot know a child. It cannot observe frustration, celebrate growth, recognize subtle growth, understand family priorities, or fully appreciate the nuances of a student’s strengths and needs. Those judgments require people. 

AI is not going away, nor should the goal be to keep it out of special education. Used responsibly, it can reduce administrative burdens, organize information, and give educators more time to focus on instruction and students. But it must remain a tool, not the decision-maker. It cannot replace professional judgment, meaningful team deliberation, educational expertise, or the relationships that lie at the heart of special education. Most importantly, it cannot assume responsibility when an IEP fails a child. Responsibility and accountability remain exactly where Congress intended: with educators, school systems, and the adults entrusted with protecting the educational rights of children with disabilities. Efficiency may be valuable, but it can never come at the expense of a student’s rights, needs, or educational progress.