The OECD.AI Policy Navigator

Our policy navigator is a living repository from more than 80 jurisdictions and organisations. Use the filters to browse initiatives and find what you are looking for.

Incorporation of Artificial Intelligence into the Electronic Administrative Records System


Added by:   OECD analyst
Added on:   30 Sep 2026
Updated by:   OECD analyst
Updated on:   30 Sep 2026

This initiative uses artificial intelligence to automatically summarise appeals submitted by citizens to the Social Security administration. It helps caseworkers quickly understand the main issues raised and classify each appeal for the appropriate handling. The system was developed to speed up processing times and manage growing workloads while ensuring citizens’ appeals are handled more efficiently.

Name in original language

Incorporación de la Inteligencia Artificial al Sistema de expediente administrativo electrónico

Initiative overview

The obligation of public administrations to resolve administrative procedures within the legal timeframe, the right of citizens to effective judicial protection, and the increasing volume of appeals against decisions issued by the General Treasury of Social Security (TGSS) have made it necessary to consider technological alternatives to expedite the processing and resolution of appeals. 

In this project, the first phase involved using AI to summarise the content of appeals received and processed through the SIMAD-SIEE (Electronic Administrative File System) application. These appeals were then classified for proper processing or redistribution to the appropriate management area. This first phase was implemented in January 2026. 

In a subsequent phase, the plan is for AI to suggest resolution proposals to the case manager regarding the issues raised in the appeal document, as well as for AI to process the information on appeals collected in our computer application. This will allow the Administration to anticipate future actions aimed at reducing administrative burdens and costs, better redistributing workloads, or preparing for future increases in workload.

Results, outcomes and impacts
Since its implementation in January 2026, the initiative has enabled the automatic summarisation and classification of appeals received by the Social Security administration, supporting caseworkers in the initial review of each file. This has reduced the time needed to understand the content of appeals and has facilitated their faster and more consistent routing to the appropriate management area. 

Results are measured through operational use of the system, review of generated summaries by caseworkers, and monitoring of processing workflows within the electronic administrative file system.

In the future, the initiative is expected to support greater efficiency in appeal handling, better workload distribution and improved capacity to manage cases.

Other relevant details

Challenges and lessons learned: One of the main challenges was ensuring that AI‑generated summaries are accurate, neutral and useful for caseworkers, given the legal relevance and diversity of appeal documents. Data protection and confidentiality were also key challenges, as the system processes sensitive administrative information. In addition, it was important to ensure that AI outputs were clearly understood as support tools, not automated decisions. These challenges were addressed by keeping human oversight in all stages, limiting AI use to summarisation and classification, and embedding privacy safeguards within the electronic administrative file system. A key lesson learned is that AI delivers most value when applied to well‑defined, repetitive tasks that reduce administrative burden without replacing professional judgement. Close collaboration between technical teams and operational units proved essential. Successful initiatives of this type require clear scope, legal and organisational safeguards, user trust.

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • AI use cases/projects in the public sector

Status:

  • Active

Start Year:

  • 2026

OECD AI Principles:

—

Other relevant urls:

—