Disability Discourse Matters: Leveraging AI for Equity in Education and Policy

Artificial intelligence is rapidly transforming the landscape of education and social science research. At the Education Collaboratory, we’ve witnessed firsthand how AI can be harnessed not just to accelerate research, but to drive meaningful change - especially in the realm of social and emotional learning (SEL) and equity. Our latest initiative, Disability Discourse Matters (DDM), exemplifies this mission by using AI to systematically track, analyze, and make public the language political leaders use when discussing people with disabilities.

What Is Disability Discourse Matters?

Disability Discourse Matters is a new public website and research tool that collects and analyzes statements made by federal, state, and local leaders about people with disabilities. Using advanced AI, these statements are scored on a four-point scale - ranging from dehumanizing to affirming - based on whether they dehumanize, criticize, deficit-frame, or affirm the value of individuals with disabilities. The results are mapped and updated regularly at disabilitydiscoursematters.org, making this data accessible to researchers, advocates, and the public.

Why Does Discourse Matter?

Language is more than words - it is a powerful tool that shapes perceptions, policies, and everyday experiences. When political leaders use deficit-based or dehumanizing language about people with disabilities, it can reinforce stereotypes, fuel discrimination, and influence policy decisions that affect millions. Conversely, affirming language can promote inclusion, opportunity, and dignity.

• Policy Impact: Deficit framing can result in policies that segregate or control, rather than empower, people with disabilities. It can also affect funding priorities, diverting resources away from infrastructure and inclusion.

• Educational Impact: In schools, deficit-based language often leads to segregation and reduced access to quality education, limiting opportunities for students with disabilities.

• Employment and Accessibility: The way disabilities are discussed influences workplace inclusion and accessibility, impacting job opportunities and economic participation.

How AI Powers DDM

The DDM project uses large language models (LLMs) to independently score political statements on their discourse about disability. This approach enables rapid, scalable, and transparent analysis of vast amounts of qualitative data - something that would be nearly impossible to achieve manually at a national scale.

The scoring rubric ranges from statements that dehumanize (score of 1) to those that fully affirm the value and potential of people with disabilities (score of 4). By mapping these results, DDM provides a real-time snapshot of how disability is discussed at the highest levels of government -and how that discourse shifts over time.

Transparency, Ethics, and Impact

DDM is committed to transparency and ethical research. The study is approved by the Yale University Institutional Review Board and pre-registered on Open Science Framework. Data is updated regularly, with the first release mapping statements from the White House Administration and upcoming expansions to Congress, the Supreme Court, and all 50 state governors.

A Model for Future Research

By integrating AI into SEL and policy research, DDM demonstrates how technology can be used thoughtfully to advance equity, inclusion, and evidence-based policy. This project not only makes critical data accessible to the public but also sets a precedent for how AI can be leveraged in other domains of education and social science.

With DDM, we aim to ensure that those words - and their impact - are visible, measurable, and actionable. We invite you to explore the data, share the findings, and join us in advocating for more inclusive and affirming discourse in education, policy, and beyond.

For more information and to view the data, visit DisabilityDiscourseMatters.org.