The Global Disinformation Policy Database.

Over the past quarter-century, governments around the world have responded to disinformation, misinformation, and related problems with statutes, decrees, and public programs. Those responses are difficult to compare. Policies that share a label differ in what they actually regulate, and policies regulating the same conduct carry different labels.

The GDPD exists to make these responses comparable. It is an open-access, cross-national dataset of national-level laws and policies addressing information disorder, coded to document both what governments do and how they define the problems they claim to address. This site is the database's public homepage. The dataset itself is available on Harvard Dataverse.

The database began at the University of Texas at Austin as a partnership between the Global Disinformation Lab and the LBJ School of Public Affairs. The team developed the GDPD for academic researchers, policymakers, journalists, and civil society organizations taking the measure of this policy area.

The GDPD covers national-level policy actions from 2000 through 2023 in jurisdictions with populations over one million. Inclusion is textual: a policy enters the database only if its text engages information disorder concepts — disinformation, misinformation, fake news, propaganda, foreign malign influence — or targets false or misleading information framed as a collective harm. Each policy is coded across 53 variables covering definitions, goals, instruments, liabilities, targeted technologies and platforms, and enforcement arrangements. The 133 governments counted here are 132 countries plus the European Union; the dataset deposit and companion paper report the country figure.

The scheme's signature variable is the Type of Targeted Information (TTI), which decomposes each policy's problem definition into four parts: the label it uses, the information quality it targets, the behavior it regulates, and the intent standard it applies. The decomposition makes conceptualization itself measurable — it is how the database can show, for instance, that a majority of “disinformation” policies specify no intent standard at all.

Coding was carried out by a team of 29 trained volunteer coders working in regional pods, with weekly calibration sessions, a shared decision-rules log, and post-hoc review of every policy by the project lead. The coding philosophy was conservative throughout: where a policy's text did not support a determination, coders recorded “unknown” rather than inferring one. The full codebook — variable definitions, coding procedures, and inclusion/exclusion rules — accompanies the companion paper and the Dataverse deposit.

A note on sources: policy records link to primary documents wherever coders located them, including government and state-media domains. Sources are presented as coded, not endorsed.

The GDPD is led by Ryan Williams, a PhD candidate at the LBJ School of Public Affairs at the University of Texas at Austin and co-founder of UT's Global Disinformation Lab. His research explores global policy responses to emerging technologies; before the PhD, he worked as a computational disinformation analyst developing tools for understanding online affinity groups.

The database would not exist without the 29 volunteer coders — graduate and undergraduate students at UT Austin — who researched, read, and coded policy documents from around the world across multiple semesters of pod-based coding work.