Fundamental Rights Impact Assessment

What Is Fundamental Rights Impact Assessment?

Learn how Article 27 of the EU AI Act requires public bodies, public-service providers, and credit or insurance scoring deployers using high-risk AI to conduct a Fundamental Rights Impact Assessment before first use. Complete a FRIA for an automated social housing benefit eligibility system, assessing impacts on non-discrimination, privacy, effective remedy, and human dignity.

What You'll Learn in Fundamental Rights Impact Assessment

• Understand when and why a FRIA is required under Article 27 of the EU AI Act
• Distinguish between a FRIA and a GDPR Data Protection Impact Assessment
• Identify all categories of affected persons for a high-risk AI system
• Assess which fundamental rights are impacted by an AI system making benefit eligibility decisions
• Define appropriate mitigation measures and monitoring plans for ongoing compliance

Fundamental Rights Impact Assessment — Training Steps

1. Article 27: Fundamental Rights Impact Assessment

   Article 27 of the EU AI Act requires certain deployers of high-risk AI systems (public bodies, private entities providing public services, and deployers of credit-scoring or life and health insurance AI) to conduct a Fundamental Rights Impact Assessment before first use. A social housing benefit system run for municipal councils is in scope: it decides eligibility for public assistance (Annex III 5(a)). A FRIA examines how the AI system might affect fundamental rights - dignity, non-discrimination, privacy, freedom of expression, and the right to an effective remedy. A FRIA differs from a GDPR Data Protection Impact Assessment (DPIA). While a DPIA focuses solely on data protection risks, a FRIA covers all fundamental rights. Both assessments may be required for the same system, and they can be conducted together.

2. Deployment Details

   An email arrives from Petra Novak, the project manager responsible for the social housing AI deployment. She provides the details Alice needs to begin the FRIA.

3. FRIA Guidance Document

   Before starting the assessment, Alice opens the FRIA guidance document linked in Petra's email. It explains each section of the assessment, the difference between a FRIA and a DPIA, and practical tips for completing each part thoroughly.

4. Open the FRIA Form

   With the guidance reviewed, Alice clicks Continue to open the FRIA checklist - five sections covering the pillars of a rights impact assessment.

5. Section 1: System Description

   The first section asks Alice to document the AI system's purpose, the people it affects, and the scope of its deployment. For this system, the affected persons include: Benefit applicants - many in vulnerable financial situations, directly affected by eligibility decisions. Case workers - who rely on the AI's recommendations to issue final decisions. Applicants' families - who depend on the housing benefits the applicant is seeking. The system makes decisions that directly affect access to housing - a fundamental need.

6. Section 1 Check

7. Section 2: Fundamental Rights Analysis

   This is the most critical section of the FRIA. Alice must assess which fundamental rights the AI system could affect: Right to non-discrimination - the AI may discriminate based on family composition, nationality, or residential address, using these as proxy variables for protected characteristics. Right to privacy - the system processes sensitive personal data including income, employment, and family information. Right to an effective remedy - applicants must be able to challenge AI-driven rejections through a clear, accessible appeal process. Right to human dignity - automated decisions about housing access affect a fundamental aspect of a person's life and dignity.

8. Section 2 Check

9. Section 3: Risk Assessment

   Alice scores each rights impact for severity and likelihood. Housing-access denial is high severity; likelihood depends on bias controls. The combined rating drives which mitigations get priority.

10. Section 3 Check