CEIAI Governance Database
Back to the database
Record STMT-WB02CA7Advisory

EDPB-EDPS Joint Opinion 5/2021 on the proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)

Opinion International lawEuropean Unionadopted

This joint opinion by the European Data Protection Board and the European Data Protection Supervisor analyzes the European Commission's proposed Artificial Intelligence Act. The document welcomes the risk-based approach and the inclusion of EU institutions within the scope, but raises serious concerns regarding data protection implications, enforcement independence, and the need for broader prohibitions on intrusive AI practices such as social scoring and mass biometric identification.

Summary and content fields are drawn from the document text and are pending editorial review.

Who issued it

European Data Protection BoardRegulator
Lead issuer
European Data Protection SupervisorRegulator
Co-issuer
Legal forceAdvisory
Under whose authorityInternational law
Judged fromfull text

Who it governs

RegulatorsNational governmentPublic sector bodiesAI developers and providersAI deployers and users

EU legislators, regulators, and policymakers should read this document to understand necessary amendments for aligning the proposed AI Act with European data protection laws.

Sectors
Cross-sector

What the document contains

What's covered

The document addresses the relationship between the proposed Artificial Intelligence Act and existing EU data protection frameworks, evaluates the risk-based approach, analyzes prohibited practices and high-risk AI systems, reviews governance and enforcement structures, and examines regulatory sandboxes, transparency, and compliance mechanisms.

What's missing

Not stated

Key takeaways
  • The EDPB and EDPS welcome the proposed Artificial Intelligence Act and its risk-based approach, but emphasize that it has profound data protection implications.
  • The opinion calls for a total ban on biometric categorization, emotion recognition AI, social scoring, and automated recognition of human features in publicly accessible spaces.
  • Compliance with existing EU data protection laws (GDPR, EUDPR, LED) and the involvement of data protection authorities as national supervisors must be explicitly mandated.
  • The conformity assessment procedures for high-risk AI systems should universally involve independent third-party checks prior to market entry.
  • The governance framework needs strengthening to guarantee the complete independence of supervisory authorities and the European Artificial Intelligence Board from political influence.
Principles it states
Primacy of fundamental rights and human dignityData protection by design and by defaultIndependence of supervisory authoritiesAccountability and transparency in AI development and deployment
Outline
· 1 Introduction
· 2 Analysis of the Key Principles of the Proposal
  - 2.1 Scope of the Proposal and relationship with the existing legal framework
  - 2.2 Risk-based approach
  - 2.3 Prohibited uses of AI
  - 2.4 High-risk AI systems
  - 2.5 Governance and European AI Board
· 3 Interaction with the Data Protection Framework
  - 3.1 Relationship of the Proposal to the existing EU data protection law
  - 3.2 Sandbox & further processing (Articles 53 and 54 of the Proposal)
  - 3.3 Transparency
  - 3.4 Processing of special categories of data & data relating to criminal offences
  - 3.5 Compliance mechanisms
· 4 Conclusion
Issued18 June 2021
Length10,683 words · 22 pages
FormatNot recorded
Scoperegional

How it scores

Pending review

Each score is backed by a quote verified against the document text. Editorial review has not yet confirmed them.

Accountability

4/5

Clear responsibility for AI outcomes

responsibilities of the various parties - user, provider, importer or distributor of an AI system - need to be clearly circumscribed and assigned

1 INTRODUCTION

screening score · 2026-10-23

Human Dignity

4/5

Respect for the inherent worth of persons

It also affects human dignity to be determined or classified by a computer as to future behavior independent of one's own free will

2.3 Prohibited uses of AI

screening score · 2026-10-23

Human Oversight

4/5

Humans remain in control of consequential decisions

The Proposal gives an important place to the notion of human oversight (Article 14) which the EDPB and the EDPS welcome.

1 INTRODUCTION

screening score · 2026-10-23

Enforcement

3/5

Consequences and compliance mechanisms with real force

the EDPS “shall act as market surveillance authority” for Union institutions, agencies and bodies that fall within the scope of the Proposal

2.5.1 Governance

screening score · 2026-10-23

Impact Assessment

4/5

Required assessment of an AI system's risks/impacts before or during use

the Proposal requires the providers of the AI system to perform a risk assessment

2.2 Risk-based approach

screening score · 2026-10-23

Topics

artificial_intelligence_act data_protection biometric_identification risk_assessment fundamental_rights regulatory_governance