Tequity in Learning

Operationalizing Equity Policies: Contextualized AI Learning Tools for Student Success Teams

San Francisco schools rely on Student Success Teams (SSTs) to address the needs of struggling students before they are referred to special education. SSTs do critical preventive work, but staff often lack up-to-date training and tools to help them design support with equity in mind. These uneven practices have contributed to long-standing racial and linguistic disproportionality in how students are identified for special education. Specifically, SFUSD has been cited for disproportionality of Black students since 2012. The researchers propose Policy-in-Practice (PiP), a new micro-learning tool that provides SSTs with realistic practice scenarios based on situations they face every day. Using generative AI and developed in close partnership with SFUSD educators, PiP turns district equity policies and school-level data into clear learning opportunities. The project aims to strengthen educators’ confidence and judgment, foster equity-aligned practices across teams, and offer the district actionable insights to refine professional learning and policy implementation.

Principal Investigators: Alfredo Artiles & Hari Subramonyam

Tequity in Breast Cancer Screening Education: Global Launch of a Cancer-Enriched Digital Breast Tomosynthesis Training Module

Breast cancer is a leading cause of cancer morbidity and mortality in many low- and middle-income countries (LMICs), particularly in Africa, where late-stage presentation and limited imaging expertise drive poor outcomes. Digital breast tomosynthesis (DBT) is the gold standard in breast cancer detection, yet high-quality DBT training and diagnostic expertise isn’t widely accessible. Building on the STELLA (STanford Electronic Learning Library & Applications) platform and the team’s successful 2025 free virtual radiology continuing medical education (CME) series for LMICs co-developed with University of Nigeria Teaching Hospital, the proposed project involves co-creating and deploying a cancer-enriched DBT training module tailored to Nigerian needs. This anonymized module will provide training that includes 300 curated DBT screening cases through a secure, web-based picture archiving and communication systems interface, supported by co-design workshops and hybrid CME sessions. The researchers hypothesize that the training module will improve diagnostic accuracy, interpretive confidence, CME access, and develop a tequity-centered framework for scaling to other LMICs.

Principal Investigator: Brittany Deshevsky

Community data sovereignty in AI x Biotechnology

Local sovereignty over data and AI models is necessary to ensure that emerging technologies serve communities and have broad benefits. Biotechnology is an emerging area using AI to improve manufacturing, agriculture, and medicine. To advance tech equity (tequity), every community must have the infrastructure to generate, steward, and benefit from their own biological data. We are creating a modular educational platform focused on community data sovereignty in AI. Collaborating with community members in Salinas (California, USA), we are developing programs and infrastructure that support data sovereignty in biotechnology.

Principal Investigator: Drew Endy

Toward a Black Geographic Approach to Education Research: A Participatory Mapping Project

This project is a community-based methodological inquiry that theorizes new ways of studying educational spaces through the interweaving of Black geographic thought and geo-spatial mapping. While educational research has increasingly attended to the role of space and place in the production of difference across race, gender, and class, few have engaged deeply with Black geographic methods and what they might reveal about schools and community-based education spaces and the experiences of racially marked youth within them. This project thinks across Black geographies in the Bay Area and Dallas, Texas, through participatory mapping workshops to explore their potential for developing educational research methods that build with the communities most impacted by educational inequities.

Principal Investigators: Zephyr Frank & Rebecca Tarlau; Research Team: Bethany Bass & Anisa Yudawanti

(Re)imagining Opportunities: GenAI-Mediated Labor Equity for Undocumented Migrant Adults

This study will investigate how GenAI can expand safe, flexible, higher-wage, and culturally sustaining pathways for undocumented migrant adults—one of the most structurally marginalized labor groups in the United States. Grounded in critical media literacy and AI literacies frameworks, this year-long qualitative study examines how undocumented migrant adults navigate AI-mediated labor opportunities while confronting systemic barriers. The team will co-design an online learning community, bringing together 8–10 undocumented adults, researchers, and vocational coaches in biweekly workshops focused on producing professional materials, exploring job opportunities, and cultivating critical understandings of algorithmic bias, digital labor precarity, and the sociopolitical harm of GenAI. This study will illuminate how GenAI can both mitigate and intensify labor inequities for people labeled as undocumented, while identifying community-driven strategies to leverage digital tools for empowerment, economic stability, and cultural affirmation. Findings will inform equity-centered and translingual AI design, adult education, and workforce development initiatives that support undocumented migrants’ full participation in an emerging (AI-mediated) labor economy.

Principal Investigators: Antero Garcia & Tairan Qiu

Engineering Neuro-Inclusion: A Participatory AI Platform for Repairing Systems, Not Students

Technology in education often focuses on tracking student deficits, which can perpetuate inequities for neurodiverse learners. Shifting from individual remediation to universal design, this project serves as a proof-of-concept for tequity in action: harnessing AI to both surface and repair inequity, thereby building truly inclusive learning communities. The project will adapt Stanford’s globally-recognized Our Voice (OV) citizen science for health equity framework—previously used in K-12 health promotion [1] and for neuro-inclusion settings like Magical Bridge [2]—to empower teachers, parents, and students to collectively assess home vs. classroom/school environments. A new AI-dashboard prototype will translate these aggregated, privacy-preserving observations into actionable Universal Design for Learning (UDL) recommendations, co-validated by participants to ensure AI enhances rather than replaces human insight.

Principal Investigator: Abby King

Leveraging Artificial Intelligence to Transform Teacher Noticing: From Deficit to Asset-Based Discourse in Mathematics Education

Mathematics teachers can have more positive impact on students by replacing deficit discourses with more positive asset-based reframing. However, deficit discourses remain pervasive, often emerging when teachers unconsciously focus on student errors as evidence of deficiency and move directly to remediation rather than building on student strengths. Deficit discourses disproportionately harm students from marginalized communities, yet go unnoticed even by equity-committed and experienced educators as they are embedded in culturally dominant assumptions that subtly shape instructional practice. This project therefore investigates whether artificial intelligence (AI) helps surface and disrupt these patterns. Specifically, we examine the potential of AI to (1) detect deficit-based linguistic features in teachers’ instructional talk, and (2) scaffold asset-based reframing through interactive, AI-mediated interventions. Thus, we seek to advance “tequity” by demonstrating how AI, when intentionally designed, helps educators surface inequitable discourse patterns and cultivate humanizing, strength-based interactions with students, ultimately promoting equitable mathematics learning.

Principal Investigators: Jennifer Osuna & John Mitchell

CRISPRkit™ for Tequity: An Integrated AI/Physical Platform to Dismantle Inequities in Biotechnology Education

To address persistent inequities in science education, the researchers propose CRISPRkit™ for Tequity: an integrated ecosystem combining the team’s proven $2 physical CRISPR education kits with a free AI-guided education app and browser-based simulations. While the low-cost kit addresses the infrastructure gap, the AI “CRISPR coach” app provides crucial, real-time instructional support, which is often lacking for Title I teachers teaching cutting-edge biotech. This project will pilot the physical and digital tools across 20+ Title I and public schools, partnering with the Amgen Biotech Experience (ABE) and LabXchange to ensure long-term integration. The researchers will use mixed-methods evaluation to investigate how the AI coach either closes or risks widening opportunity gaps, ensuring this technology is an intentional force for equity.

Principal Investigator: Stanley Qi

AI-mediated growth mindset and culture of learning intervention with high-school students and teachers in mathematics

Students from vulnerable socioeconomic backgrounds in Latin America face high rates of absenteeism and dropout and significant inequalities in mathematics achievement, with girls particularly affected. These disparities limit long-term opportunities and underscore the need for interventions that enhance motivation, persistence, and learning in mathematics classrooms. This project integrates growth-mindset theory with AI-mediated dialogic learning for students and teachers, aiming to advance educational equity in mathematics. Students engage with a virtual character with a fixed mindset in structured dialogues that foster empathy, argumentation, and reflection. Teachers participate in an 8-week professional learning program focused on growth mindset, agency, and formative feedback practices. The one-year proof-of-concept involves five mathematics classes in both Colombia and in Uruguay and emphasizes diverse socioeconomic contexts. By combining scalable AI technology with evidence-based interventions, this project aims to address persistent inequities, informs high-fidelity intervention design, and generates insights for inclusive, technology-enhanced educational practices globally.

Principal Investigator: Micaela Bonilla

Tequity in Motion: Co-Designing AI-Enhanced Education for Body Image, Nutrition, and Empathy Across Athletic and K–12 Learning Communities

This project explores how AI and digital learning tools can advance tequity by promoting empathy, positive body image, and reflective dialogue about wellbeing in collegiate and youth athletic settings. Although many collegiate athletic programs offer access to wellness resources, student-athletes often face barriers to seeking support, including stigma surrounding mental health, limited time, and performance-oriented cultural norms. Similar challenges can arise for adolescents participating in school sports and youth sports clubs, where opportunities for structured discussion of body image and wellbeing may be limited. Using a community-centered, mixed-methods approach, the project will co-develop and pilot Tequity in Motion, an educational intervention designed to encourage empathy, self-reflection, and dialogue around body image and wellbeing. The study aims to identify barriers to inclusive wellness education and to inform the development of scalable learning tools that can be adapted across athletic and educational settings.

Principal Investigator: Koto Imahori

NumberNestAI: Breaking Math Achievement Gaps Through SMS-Based, Culturally-Adaptive Parent Empowerment

Many parents experience math anxiety when helping their children, perpetuating achievement gaps that emerge by age 5, before formal schooling begins. This disproportionately affects households that lack access to existing digital learning tools. This project transforms everyday family moments into culturally-responsive math learning opportunities through an AI system accessible via SMS. By leveraging multi-agent architecture with systems trained on diverse parenting practices, NumberNest generates personalized, play-based activities that respect each family’s culture and routines. This project will conduct user research with 30+ diverse families to refine the team’s equity-first approach, pilot SMS deployment with 50 families across income and language groups, and measure dual-generation impact on both child math identity and parent confidence. Unlike solutions that digitize Western pedagogy, NumberNest reimagines AI as a bridge between cultural wisdom and academic standards. Through this project, technology repairs rather than reinforces inequities, making high-quality math support accessible to the 4 million+ preschoolers whose families currently lack resources for early math enrichment.

Principal Investigator: Xi Jia Zhou