The EU Commission proposes expanding Europol’s powers by loosening procedural safeguards, enabling mass data collection and extensive use of AI in policing.

A June 2026 EU proposal would deregulate Europol, broaden data gathering, and increase AI deployment, critics arguing it would weaken privacy protections and safeguards.

Experts from EDPS and other groups warn that AI systems must be auditable and linked to known legal standards, emphasizing transparency and explainability.

The plan would create private testing environments for tech firms and reduce transparency requirements, raising concerns about misinformation, discrimination, and limited redress for individuals.

The draft regulation would reclassify compliance burdens as administrative, accelerating data processing and enabling near-anonymous searches without tying data to specific crimes or suspects.

Private-sector testing in Europol’s Innovation Lab could speed up deployment of unproven AI tools, with cited incidents and international examples illustrating governance and security risks.

Critics argue Europol has historically operated covert data-processing systems to bypass safeguards, and reform could normalize or expand such practices.

Europol would become a central hub for information from private actors and gain access to funding and participation in Horizon Europe, signaling broad capability expansion in digital policing tools.

The plan envisions closer private-sector collaboration, allowing data exchange with private actors, tapping large funding streams (including Horizon Europe’s multi-year totals) to develop digital surveillance tools.

Existing exemptions from AI Act monitoring for Europol raise risks of privacy erosion, potential abuses, and discriminatory impacts on migrants and asylum seekers.

Statewatch and EU data-protection authorities challenge the plan, noting conflicts with EU Court rulings on traceability and privacy and warning it could weaken fundamental rights under AI governance.

The proposal would allow Europol to search for and store data about people not linked to an investigation, reducing safeguards that ensure data relevance and proportionality.