Home›Research

Working
Literature
Review

Focus AI, Education & the Black Community
Grade Range Grades 5–9
Methodology Community-Based Participatory Research
Status Active — Updated Ongoing

This is a living document. I am sharing my research process in real time — what I am reading, what I am finding, and where the gaps are. This page grows as the work grows. Come back often.

Core Research Question

How is the adoption of AI tools in grades 5–9 affecting the educational experiences and outcomes of Black students, and what community-based infrastructure is needed to ensure equitable implementation?

Five Themes.
One Connected Story.

Read Full Document
01

Racial Bias in AI Educational Tools

The foundational problem: AI tools in education carry racial bias because they reflect the biases of their designers. The tech industry is 92.6% non-Black. The tools being built for Black students are rarely built by people who understand their experience — and that gap shows up in the classroom.

AI-powered tools may present incomplete accounts of Black history, misinterpret cultural references in student essays, and fail to accurately identify which students need support. These are not edge cases. They are the predictable output of a design process that does not center Black students.

The Research Gap

No community-level study has examined how bias in AI tools specifically manifests in grades 5–9 classrooms serving majority-Black populations in mid-size California cities. This is EduBlck Institute's entry point.

Key Sources

McKinsey & Co. (2023)

EdTech Equity Project (2023–24)

Stanford Center for Racial Justice (2024)

Frontiers in Education (2024)

02

The Digital Divide and AI Access

Before bias in the classroom can be addressed, there is a prior problem: unequal access. A 2023 Pew Research Center survey found that 72% of White teens had heard of ChatGPT, compared to 56% of Black teens. The awareness gap begins in the home — before any teacher deploys a tool.

In California specifically, 35% of low-income households still lacked reliable internet in 2021. In Pittsburg, CA — the home base of this research — broadband equity and device access are live community issues, not abstract national statistics.

The Research Gap

No study has measured AI awareness and access specifically within the Pittsburg, CA community or comparable Bay Area working-class Black communities. This is primary data waiting to be gathered.

Key Sources

Pew Research Center (2023)

PPIC California (2021)

Stanford Center for Racial Justice (2024)

03

Teacher Training and Institutional Failure

As of fall 2024, only 47% of teachers reported receiving any training on AI — and that training was typically a one-time event, not ongoing professional development (RAND, 2025). More than half of American teachers are navigating AI-integrated classrooms with no institutional guidance.

For under-resourced districts serving predominantly Black students, this is compounded: Black students receive tools not designed for them, deployed by teachers not trained to use them, in schools that cannot afford to address either problem.

The Research Gap

No research has gathered primary data specifically from Black educators on their experience of AI professional development inadequacy and the direct effect on Black students in their classrooms.

Key Sources

RAND Corporation (2025)

U.S. Dept. of Education (2023)

Stanford Center for Racial Justice (2024)

04

The Achievement Gap at the Middle School Level

Grades 5–9 is the developmental window where academic identity forms and the achievement gap is most consequential. As of spring 2025, only 39.3% of 8th graders in majority-Black schools met grade-level reading standards. In mathematics — this researcher's content area — students remain 1 to 14 percentage points below pre-pandemic 2019 levels.

The achievement gap is now 30% larger than it was 35 years ago — driven by the dismantling of equity-focused policies from the 1960s and 70s. The pandemic accelerated a pre-existing crisis. AI is arriving into this context.

The Research Gap

No study has examined how AI math remediation tools perform for Black students in grades 5–9 in California, or how those students and teachers experience these tools culturally and academically.

Key Sources

EdWeek (2025)

Darling-Hammond / LPI (2024)

Hechinger Report (2024)

EdTrust (2024)

05

Frameworks for Equity-Centered Research

This research is grounded in three interlocking frameworks. Culturally Responsive Pedagogy (Ladson-Billings) holds that student cultural identity is an asset to be centered, not a variable to be managed. Community-Based Participatory Research (CBPR) ensures that the community most affected by a problem is an active participant in researching it. Carter G. Woodson's foundational work reminds us that this is not a new problem — only a new mechanism.

Together these frameworks define how EduBlck Institute collects data, interprets findings, and builds community trust.

The Research Gap

CBPR methodology has not been applied to AI-in-education research with an insider researcher embedded in a specific Black community in California. That combination is EduBlck Institute's original contribution.

Key Sources

Ladson-Billings (CRP)

Woodson (1933)

CBPR Framework

ScienceDirect (2025)

Why This Research Matters

What exists — and what doesn't.

What the literature already covers

  • National-level data on AI bias in educational tools
  • Aggregate achievement gap statistics for Black students
  • Policy recommendations from government and think tanks
  • Theoretical frameworks for equity-centered AI in education
  • Surveys of teacher AI training nationally

What EduBlck Institute contributes

  • Hyperlocal research grounded in Pittsburg, CA
  • Primary data from Black educators on AI in grades 5–9
  • CBPR methodology with an inside researcher
  • A math-specific lens grounded in subject-matter expertise
  • Community trust infrastructure no outside researcher can replicate

What I'm Reading

Foundational Texts

  • Woodson — The Mis-Education of the Negro (1933)
  • Ladson-Billings — Dreamkeepers
  • Noble — Algorithms of Oppression (2018)

AI & Education Equity

  • U.S. Dept. of Education AI Report (2023)
  • EdTrust — Promise and Peril of AI (2024)
  • Stanford Center for Racial Justice (2024)
  • Word In Black Series — AI in Schools (2024)
  • EdTech Equity Project Toolkit (2023–24)

Achievement Gap & Middle School

  • Darling-Hammond — AERA Address (2024)
  • Hechinger Report — Achievement Gap Since Brown (2024)
  • RAND — Teacher AI Training Report (2025)
  • EdWeek — Post-Pandemic Achievement (2025)

Methodology

  • Israel et al. — CBPR for Health
  • Minkler & Wallerstein — CBPR Framework
  • ScienceDirect — Inclusive AI in K-12 (2025)

Read the full working document.

The complete literature review includes full source citations, researcher notes, field observations, and a detailed breakdown of each theme. It is updated as the research develops — check back for new additions.

Open Full Document View only · Updated ongoing · Last updated 2025