Meredith Bouvier

Meredith Bouvier

AI, Children & Online Safety · Computational Research
Platform Governance & Algorithmic Accountability

About

I am a computational researcher studying how AI and platform systems shape the lives of children — in classrooms, at home, in games, and in their interactions with increasingly capable AI tools. I use machine-learning-assisted text analysis, topic modeling, digital trace data, and survey research to understand how these systems behave in practice, how they can be measured, and how they can be made safer and more accountable to the people most affected by them.

Much of my work centers on a question that matters more as AI systems reach younger users: how do we evaluate and govern technologies built for or used by children, so that they support healthy development rather than exploit attention, data, or trust? I am especially interested in age-appropriate design, the gap between what platforms claim and what they do, and surfacing the knowledge of children, parents, and educators who are rarely consulted in the systems making decisions about their digital lives.

I am Co-founder and Chief Research Officer of the Classroom Tech Transparency Project, where I build privacy-preserving tools and computational methods for understanding how digital systems are used by children across schools and households. I recently completed my PhD at Penn State.

My background is non-linear in ways I consider an asset. I hold mathematics degrees from Duke University and Columbia Teachers College, and before turning to computational research I taught mathematics to refugee and immigrant students in New York City, lived and taught abroad, and held research positions at the World Bank, UNESCO IIEP, OECD, and the Global Partnership for Education.

Outside of work, I swim competitively, bike across countries, and dig in the dirt with my two bilingual boys.

My methods are computational: Python and SQL pipelines, LLM-assisted analysis, BERTopic, and digital trace data. CV available upon request at merrybouv@proton.me.

Selected Publications

Some additional publications are listed on Google Scholar. For a full list including policy reports, working papers, and grey literature, request my CV at merrybouv@proton.me.

Current Projects

Classroom Tech Transparency Project (CTTP)

in beta

CTTP is a privacy-first research and tools organization that helps parents and caregivers understand the technology used in their children's classrooms — what it surfaces, how it behaves, and how its use compares across schools and households. Co-founded with parent advocate Joanna Houston, CTTP has attracted interest from multiple national news outlets ahead of its public launch.

  • ClickBelow (live) — uses Google Takeout exports, combined with school bell schedules, to surface the YouTube videos a child's account accessed during instructional time. Enables evidence-based conversations with teachers and schools about platform use and content exposure.
  • EdTech Concern Explorer (in development) — a searchable interface into a curated 10-year corpus of Reddit discourse (2014–2024) from parents, teachers, and students. Search by platform or concern type and read what real families, educators, and students are saying about technology in school.
  • What Parents & Teachers Report (live) — a participatory data platform built from CTTP's original survey research. Parents and educators can contribute their experiences and see how their responses compare to the growing aggregate.
  • FERPA Request Tracker (in development) — generates customized FERPA request letters naming specific EdTech platforms, tracks the legally required 45-day district response window, and sends follow-up and escalation letters automatically.

Co-founded with Joanna Houston · classtechtransparency.org

Children Online: A Digital Discourse Corpus

in development

A computational corpus mapping public discourse on children's digital environments beyond the classroom. Drawing on parenting, gaming, and child-safety communities, it uses topic modeling and large-scale text analysis to surface how the public understands online risks to children — age-inappropriate content, gaming platform design, AI companions, online predation, and the governance gaps that allow harm to persist. Themes are derived empirically from the data rather than imposed in advance.

A companion bibliometric study examines whether the research literature on children and gaming centers platform accountability or deflects responsibility onto individual children and families.

Independent research

AI in Education: Global Competency Frameworks

in progress

Comparative computational analysis of national AI education policy documents across approximately 30 countries. Uses Claude API structured extraction to identify competency statements, followed by BERTopic thematic analysis and cross-country comparison.

With a distinguished professor of education policy, Penn State

Apprenti

prototype

An offline-first, AI-powered career exploration tool for students. Apprenti generates personalized, skill-based missions that connect classroom learning to real-world work — without tracking, without data collection, and without requiring school infrastructure.

Independent research prototype

Contact

Interested in research collaborations or applied AI and child-safety work? Reach out below.