Skip to content
Submit incident
AII-1000-41527
AllegedThe record is a lawsuit, complaint or accusation that has not been decided.

Greece uses AI to alter 54,500 public employee evaluations without legal basis

Greece's Ministry of the Interior reportedly used an AI tool to reprocess performance evaluations of 54,500 public employees in 2023, reclassifying most of them downward without statutory authorization or transparency safeguards. Advocacy groups filed complaints with the National Transparency Authority, alleging violation of Greek AI governance law.

7 August 2026 · Ministry of the Interior · Governance, accountability and labour
7 August 2026
AII-1000-415272 sources ↓
Quick facts
Category
Governance, accountability and labour
Country
Greece
Source
NEWS 24/7
Date
7 August 2026
Drafted by
Raidu Incident Detection

What happened

The Greek Ministry of the Interior processed data on 54,500 public employees using an AI tool to reassess their 2023 performance evaluations, after they had initially been rated as high-performing. Through this processing, approximately 50,000 of those evaluations were rejected, reducing the final count of high-performing employees to fewer than 4,500, with 1,329 employees reclassified as low-performing.

Reported impact

The reclassifications affected eligibility for career advancement, as a high-performance rating is a prerequisite for entry into a proposed register of supervisors.

Index analysis

Under Greek law 4961/2022, the government may use AI systems for decisions affecting citizens' or workers' rights only when explicitly permitted by a specific legal provision and accompanied by defined safeguards. No such specific legal provision exists for public employee evaluations. The Ministry of the Interior did not publish a required impact assessment or transparency information mandated by law, and it is unclear whether employees were clearly informed how the AI system affected their assessments.

The EU AI Act classifies such high-risk applications as involving algorithmic bias, lack of transparency, and accountability failures.

Governance controls

Homo Digitalis, the Institute of Regulatory Policy, and a public employee filed coordinated complaints with the National Transparency Authority on 3 July, seeking investigation of whether the Ministry complied with legal obligations for AI system use in public employee evaluations.

Classification

Category
Governance, accountability and labour
Industry
Government, security and defence
Country
Greece
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Reputational
Harmed
Workers
Failure modes
Biased model
Consequences
Regulatory investigation
AI systems
AI tool for public employee evaluation
Regulations
EU AI Act (cited)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

NEWS 24/7
Καταγγελία στην ΕΑΔ: Παράνομη χρήση AI στην αξιολόγηση 54.500 δημοσίων υπαλλήλων
7 August 2026
Η Εφημερίδα των Συντακτών
Δημόσιοι υπάλληλοι: Καταγγελία για την αξιολόγηση μέσω Τεχνητής Νοημοσύνης
10 August 2026