Data & Society Research Institute

Data & Society Research Institute is a New York City-based independent nonprofit research and policy institute founded in 2014 to study the social, political, and economic implications of data-centric technologies, automation, and artificial intelligence. Its core operating premise — that technology policy must be grounded in empirical research, account for real-world impacts, and serve the public — positions the organization as movement infrastructure: producing the evidence base that civil society, legislative staff, and regulatory bodies carry into binding accountability mechanisms.

Founding and structure

Data & Society was founded by danah boyd, a researcher and partner at Microsoft Research who brought seed funding from Microsoft to establish the institute in 2014. boyd served as president from the founding through approximately 2022, building the organization from a small fellows program into a multi-program research institute with staff across research, policy engagement, communications, and development. The organization is currently led by Executive Director Janet Haven, who previously served at the Open Society Foundations. MacArthur Foundation grant records document 14 grants totalling $4.76 million from 2014 through 2024 and name the institute's principal research programs, confirming the trajectory from a digital-media-and-learning focus at founding to an AI-and-automation focus from approximately 2018 onward. Data & Society operates from Chelsea, Manhattan, with an international network of affiliated researchers.

Research programs

The MacArthur grantee record identifies five principal research lines: AI on the Ground, which assesses AI system impacts to inform design and governance; Labor Futures, which uses ethnographic research to examine workforce disruptions from technological development; Trustworthy Infrastructure, which works with marginalized groups on online trust-building; the Media Manipulation Initiative, which examined platform governance and algorithmic issues; and Health, Data and Equity, which explored data equity in healthcare contexts. The Algorithms and Publics program developed empirical, participatory algorithmic impact assessment methodologies in support of algorithmic governance. A 2026 program, Building Civic Strength for an AI Era, focuses on AI civics education and public capacity-building.

Policy engagement and the Algorithmic Accountability Act

Data & Society's fellow-to-policy pipeline produced one of the most direct connections between civil-society AI research and federal legislation in the United States. In 2018, fellow Mutale Nkonde, working as a tech policy advisor for Congresswoman Yvette Clarke, organized a briefing for the Congressional Progressive Caucus built around the institute's Algorithmic Accountability: A Primer report, which Nkonde had led at Data & Society. The briefing provided Congressional staff with accessible framing on fairness, bias, opacity, and auditing in automated systems. In 2019, Clarke and colleagues introduced the Algorithmic Accountability Act, which would require companies to conduct and remediate impact assessments for algorithms producing consequential decisions — directly translating the institute's audit-assessment research agenda into proposed legislation. Joan Donovan, who led the Media Manipulation Initiative, testified before the House Committee on Oversight and Reform in October 2020 on platform accountability. The Office of Management and Budget's draft memo on federal agencies' use of AI subsequently highlighted two methodologies the institute had long advocated — algorithmic impact assessments and public participation in the AI life cycle — as recommended practice for federal agencies.

Public Technology Leadership Collaborative

The Public Technology Leadership Collaborative, a Data & Society initiative documented in MacArthur Foundation grant records, functions as a peer-learning collective for government officials and academic researchers studying data-centric technologies. The Collaborative creates a structured channel between the institute's empirical research and practitioners inside public-sector institutions — a non-legislative form of policy engagement that complements the congressional-testimony and legislative-advocacy work. It operates alongside the institute's public-facing Points blog and research library, which makes findings accessible to journalists, advocates, and civil-society organizations outside the researcher community.

Role in the movement

Data & Society operates structurally as a knowledge-production organization whose outputs — primers, working papers, testimony, fellows alumni who move into policy roles — circulate through the movement rather than anchoring a single campaign. Its closest analogues in the corpus are the AI Now Institute and Upturn, both of which pursue similar research-to-policy models on algorithmic accountability; Algorithmic Justice League overlaps on the empirical-audit research form but is primarily an organizing body where Data & Society is primarily a research and policy institute. The institute's most durable movement contribution is the set of shared analytic frameworks and vocabulary — algorithmic accountability, automated decision systems, data equity — that grassroots advocates, litigators, and legislative staff now use without needing to coin them from scratch.