Jump to content

European Digital Rights

From AI Law Wiki

AI-generated text. This page was generated using artificial intelligence.

European Digital Rights (EDRi) is a network of nongovernmental organizations working on human rights in the digital environment. Its AI-policy work addresses surveillance, discriminatory decision-making, access to remedies and the obligations of companies and public authorities. It combines advocacy at European institutions with the work of national and regional member organizations.[1]

Common name: EDRi.
Legal form: International nonprofit association (AISBL) based in Brussels.[2]
Policy focus: Human rights in digital technologies.[1]
Official site: edri.org
Last source review: September 12, 2026

Institutional role

EDRi describes its priorities in terms of privacy and freedom from surveillance, democratic participation and an open internet. It seeks to connect debates about technical regulation with the experience of people affected by data-driven systems. Its network structure matters for attribution: a member's publication is not automatically an EDRi institutional position, and a coalition statement represents its named participants rather than every organization in the network.[1]

Its AI document pool collects organizational statements, consultation responses and resources from member organizations. The collection provides a record of interventions during the development of European AI legislation, including recommendations that changed as negotiations progressed. The dates and authors of individual documents are therefore more informative than treating the collection as a single, current description of the law.[3]

EU AI Act negotiations

In July 2023, EDRi joined a statement by more than 150 civil-society organizations calling for stronger rights protections during the EU AI Act trilogue. The participants sought fundamental-rights impact assessments, public information about high-risk deployments, accessibility requirements and effective complaints and redress. They opposed provisions that would allow providers to avoid high-risk obligations through their own classification of systems.[4]

The statement also advocated prohibitions on particular surveillance and predictive uses and stronger protections for people affected by migration-control systems. These were negotiating demands. They should not be read as a checklist of provisions ultimately enacted, nor as proof that a final provision resulted from this intervention. The document illustrates EDRi's emphasis on rights enforceability alongside technical risk-management duties.[4]

Predictive policing and criminal justice

In March 2022, EDRi and Fair Trials coordinated a call by more than 40 organizations for an AI Act prohibition on predictive systems in policing and criminal justice. The statement argued that systems drawing on historical policing data can reproduce patterns of discriminatory enforcement while presenting their outputs as objective assessments.[5]

The coalition connected these concerns to racial profiling, the presumption of innocence and fair-trial rights. Its argument extended beyond improving model accuracy: it questioned whether predicting criminality from personal characteristics or associations is compatible with those rights. This was a policy position advanced during legislative negotiations, not a judicial finding that every predictive system violates European law. The distinction helps explain why EDRi's proposals sometimes seek prohibitions rather than audits or disclosure alone.[5]

Funding and independence

EDRi publishes foundation, corporate and other funding information and identifies itself as an international nonprofit association in Brussels. Its 2026 list includes the Ford Foundation, Luminate, Open Society Foundations and the European AI and Society Fund, among other supporters. It also receives membership fees and corporate donations.[2]

Its published policy caps total corporate donations at 30 percent of the yearly budget and a single corporate donation at 7.5 percent of the previous year's total income. EDRi states that receiving a donation does not imply endorsement of a donor's policies. These are the organization's disclosed safeguards; they do not establish which donor, if any, supported a particular AI publication. Funding relationships and policy agreement should be assessed separately.[2]

Selected contributions

  • Joint call on predictive policing and criminal justice (March 2022).[5]
  • Civil-society demands for AI Act trilogue negotiations (July 2023).[4]
  • AI and fundamental-rights document pool, distinguishing EDRi publications from member resources.[3]

References