evidoria

← Back to browse

Good practice Imported

AI-Assisted Legal Review at the Hellenic Cadastre — Cutting a National Backlog by Up to 90%

Greece · Athens · See the Greece profile · See the Athens profile

Evidence: Observational / pre–post Top 14% 80/100 · Ask Evidence Copilot about this practice

Greece's Hellenic Cadastre became the first Greek public authority to use AI in administrative decision support, reading contracts and flagging missing legal elements for staff. Backlogs fell 80% in Athens and 90% in Thessaloniki as AI-assisted acts reached 310,000.

310,000 acts
Registrable acts completed with AI assistance (by mid-2026)
~1.1 million acts
Total registrable acts completed since 2024 (2024 to Q1 2026)
80 %
Backlog reduction, Athens cadastral office (2024-2026)
90 %
Backlog reduction, Thessaloniki cadastral office (2024-2026)
79,200 acts
Single-month record of acts completed (March 2026)

Details

Maturity
Scaling
Promoter
Hellenic Cadastre (Ελληνικό Κτηματολόγιο), Ministry of Digital Governance
Period
2024–2026
Region (NUTS)
EL30
Keywords
land registry, digital government, legal tech, back-office automation

Context

The Hellenic Cadastre, Greece's public body responsible for land registration, became what its president Stelios Sakaretsos called the country's first public authority to use artificial intelligence in an administrative decision-support role.

Objectives

The tool, presented by Digital Governance Minister Dimitris Papastergiou and Deputy Minister Konstantinos Kyranakis, reads the natural-language text of a submitted deed, identifies the type of legal act, and checks it against required elements such as tax declarations and ENFIA property-tax filings.

Activities

The system hands the reviewing officer a recommendation to approve or reject each submitted deed, with the final decision staying with the human official rather than the AI.

Results

By mid-2026 the Cadastre reported that 310,000 registrable acts had been completed with AI assistance out of roughly 1.1 million completed since 2024. Backlogs of pending cases fell by a reported 80% at the Athens cadastral office and 90% at Thessaloniki's, figures cited in a Greek Court of Audit report and repeated across independent outlets.

Conclusions

The figures come mostly from the Cadastre and government ministries rather than an independent technical audit of the AI system itself, and no error or false-recommendation rate has been published; the acceleration also coincides with a broader digitisation push (600 million archive pages scanned) that is not solely attributable to AI.

Implementation
Indicative cost
Medium (€50k–€500k)
Time to results
Medium (1–3 years) — Deployed nationwide across cadastral offices from 2024, with results reported through the first quarter of 2026.
Staffing & skills
Reviewing officers who retain the final approve/reject decision on every deed, Ministry of Digital Governance leadership (Minister Dimitris Papastergiou, Deputy Minister Konstantinos Kyranakis)

Conditions for success

  • Keeping the AI to a decision-support/recommendation role, with the final decision retained by a human official
  • Integration with existing tax and property records (ENFIA filings) for automated cross-checking
  • Part of a broader digitisation push (600 million archive pages scanned) that supports the same workflow

Common failure modes

  • No error or false-recommendation rate has been published for the AI's suggestions.
  • The acceleration coincides with a broader digitisation push, so the gains are not solely attributable to the AI component.

Where it fits

Governance type
national land registry authority
Scale
national
Income level
high-income

Commonly funded by

National / regional programmes Digital Europe Programme

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful