CHL-003

SUSESO ML-Based Medical Insurance Claims Processing

Download PDF
Chile Latin America & Caribbean High income Design & Development Phase Confirmed

Superintendencia de Seguridad Social (SUSESO)

At a Glance

What it does Classification — Operational and process automation
Who runs it Superintendencia de Seguridad Social (SUSESO)
Programme SUSESO Medical Insurance Claims Processing System
Confidence Confirmed
Deployment Status Design & Development Phase
Key Risks Governance and institutional oversight risks
Key Outcomes No published performance metrics or outcomes data.
Source Quality 4 sources — Report (multilateral / development partner), News article / media, Government website / press release

Chile's Superintendencia de Seguridad Social (SUSESO), the social security and medical insurance supervisory agency, is developing and deploying machine learning models to assist its claims processing staff in handling a high volume of medical insurance claims. SUSESO's claims processing division has approximately 80 staff, some of whom are doctors and health specialists, who process approximately 200,000 medical insurance claims per year.

The World Privacy Forum (WPF) documented two distinct ML projects at SUSESO in its November 2024 report. First, a gradient boosting model to optimise management of medical leave claims, which was in the development phase as of that reporting. Second, classification trees for occupational mental health claims, which was already in operational use and undergoing a bias audit. Both models are managed by Rodrigo Moya, who heads SUSESO's Digital Transformation, Innovation and Project Unit in the Technology and Operations Department.

The governance context is significant and well-documented. Chile has developed an 'AI bidding template' through its ChileCompra public procurement system -- a government procurement tool that includes responsible AI criteria for assessing AI vendors. SUSESO's experience implementing this tool highlighted tensions between traditional procurement evaluation (cost, technical capability) and responsible AI assessment (fairness, discrimination risk, transparency). Kate Kaye, Deputy Director of WPF, published a follow-up article in April 2025 examining how Moya changed his perspective on the AI governance tool through the process of applying it to the medical insurance claims models.

When medical leave payments are denied, those decisions can and often do have life-changing effects on claimants, making the governance of the AI component a high-stakes question. The fact that SUSESO is conducting a bias audit on the operational classification tree model indicates awareness of these risks.

SUSESO is one of seven public institutions piloting GobLab UAI's (Universidad Adolfo Ibanez) Ethical Algorithms tools, as confirmed by an official SUSESO press release. The ethical tools applied to SUSESO's ML projects include an Algorithmic Impact Assessment, an Algorithmic Transparency Report Card, and a Bias and Equity Measurement framework. The audit process covers data sources, variables, parameters, bias, equity, transparency, accuracy, counterfactual analyses, and explainability mechanisms. Superintendent Pamela Gana and project manager Rodrigo Moya presented SUSESO's AI initiative at the second State Modernization Experience Exchange session (SIEME) organised by Chile's Ministry of Finance in October 2024.

Reporting by Diario Financiero (November 2024) provided additional detail on the scope of SUSESO's AI ambitions. The agency is developing six AI-enabled projects with a combined investment of approximately CLP $550 million, funded through State Modernization Secretariat grants and the Research and Development Agency. The flagship initiative is the predictive model to optimise medical license dispute resolutions, expected to integrate by mid-2025. A second project targets fraudulent medical licensing, described by Superintendent Gana as a tool to identify, investigate and sanction health professionals potentially issuing unfounded documents. Additional projects use NLP for claim classification and data analytics for systemic failure detection. Full implementation across all six initiatives is anticipated by Q1 2026.

SUSESO processes approximately 180,000 complaints annually, with medical licenses comprising approximately 70 percent of the caseload. The agency's engagement with responsible AI governance from the earliest stages of development -- including formal bias auditing, algorithmic impact assessment, and transparent procurement through ChileCompra -- makes this case notable as an example of proactive AI governance in a Latin American social protection institution.

Classifications follow the DCI AI Hub Taxonomy. Hover over field labels for definitions.

Social Protection Functions

Implementation/delivery chain
Assessment of needs/conditions + enrolment primary
SP Pillar (Primary) The social protection branch: social assistance, social insurance, or labour market programmes. Social insurance
Programme Name SUSESO Medical Insurance Claims Processing System
Programme Type The type of social protection programme, classified under social assistance, social insurance, or labour market programmes. View in glossary Work injury and occupational insurance
System Level Where in the social protection system the AI is applied: policy level, programme design, or implementation/delivery chain. View in glossary Implementation/delivery chain
Automation Subtype For operational automation cases: (a) document processing and generative staff assistance, or (b) workload and resource forecasting. (a) Document processing and generative staff assistance
Programme Description ML-assisted medical insurance claims processing system used by SUSESO to help approximately 80 staff evaluate approximately 200,000 annual claims related to medical leave wages and occupational mental health costs.
Implementation Type How the AI output is produced: Classical ML, Deep learning, Foundation model, or Hybrid. Affects validation, compute requirements, and governance profile. View in glossary Classical ML
Lifecycle Stage Current stage in the AI lifecycle, from problem identification through to monitoring, maintenance and decommissioning. View in glossary Integration and Deployment
Model Provenance Origin of the AI model: developed in-house, adapted from open-source, commercial/proprietary, or accessed via third-party API. View in glossary Not documented
Compute Environment Where the AI system runs: on-premise, government cloud, commercial cloud, or edge/device. View in glossary Not documented
Sovereignty Quadrant Classification of data and compute sovereignty: I (Sovereign), II (Federated/Hybrid), III (Cloud with safeguards), or IV (Shared Innovation Zone). View in glossary Not assessed
Data Residency Where the data used by the AI system is stored: domestic, regional, or international. View in glossary Not documented
Cross-Border Transfer Whether data crosses national borders, and if so, whether documented safeguards are in place. View in glossary Not documented
Decision Criticality The rights impact of the decision the AI supports. High criticality requires HITL oversight; moderate requires HOTL; low may operate HOOTL. View in glossary High
Human Oversight Type Level of human involvement: Human-in-the-Loop (active review), Human-on-the-Loop (monitoring), or Human-out-of-the-Loop (periodic audit). View in glossary HITL
Development Process Whether the AI system was developed fully in-house, through a mix of in-house and third-party, or fully by an external provider. View in glossary Not documented
Highest Risk Category The most significant structural risk source identified: data, model, operational, governance, or market/sovereignty risks. View in glossary Governance and institutional oversight risks
Risk Assessment Status Whether a formal risk assessment, informal assessment, or independent audit has been conducted for this system. Formal assessment
Documented Risk Events WPF reporting documented tensions between traditional procurement criteria (vendor cost) and responsible AI criteria (discrimination, bias). Chile's AI bidding template implementation at SUSESO highlighted that minute decisions in AI governance tool design directly affect whether governance works in practice.

Risk Dimensions

Governance and institutional oversight risks

Impact Dimensions

Autonomy, human dignity and due process
Equality, non-discrimination, fairness and inclusion
  • Bias audit
  • DPIA/AIA conducted
  • Human oversight protocol
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
Administrative data from other sectorsSensitiveSingle source (no linkage)Currently available and usedMedical insurance claims data including medical leave documentation and occupational mental health records. Processed by approximately 80 staff including doctors and health specialists.

Kaye, K. (2024) 'AI Governance on the Ground: Chile's Social Security and Medical Insurance Agency Grapples with Balancing New Responsible AI Criteria and Vendor Cost', World Privacy Forum, 1 November. Available at: https://worldprivacyforum.org/posts/ai-governance-on-the-ground-chiles-social-security-and-medical-insurance-agency-grapples-with-balancing-new-responsible-ai-criteria-and-vendor-cost/ (Accessed: 30 March 2026).

View source Report (multilateral / development partner)

Kaye, K. (2025) 'Why Rodrigo Moya Changed His Mind About Chile's AI Governance Tool for Assessing a Medical Insurance Claims AI Model', World Privacy Forum, 29 April. Available at: https://worldprivacyforum.org/posts/why-rodrigo-moya-changed-his-mind-about-chiles-ai-governance-tool-for-assessing-a-medical-insurance-claims-ai-model/ (Accessed: 30 March 2026).

View source Report (multilateral / development partner)

Zecchetto, M. (2024) 'La estrategia de la Suseso para identificar licencias medicas falsas y mejorar respuesta a reclamos con IA', Diario Financiero, 12 November. Available at: https://www.df.cl/df-lab/transformacion-digital/la-estrategia-de-la-suseso-para-identificarlicencias-medicas-falsas-y (Accessed: 30 March 2026).

View source News article / media

SUSESO (2024) 'Suseso es parte de las instituciones publicas que participan del piloto para el uso responsable de inteligencia artificial', SUSESO. Available at: https://www.suseso.cl/605/w3-article-731254.html (Accessed: 30 March 2026).

View source Government website / press release
Deployment Status How far the system has progressed into real-world operational use, from concept/exploration through to scaled and institutionalised. View in glossary Design & Development Phase
Scale / Coverage The scale and geographic or population coverage of the deployment. Approximately 200,000 medical insurance claims processed annually by 80 staff with ML assistance
Funding Source The source(s) of funding for the AI system development and deployment. SUSESO operational budget (Chilean government)
Technical Partners External technology vendors, academic partners, or development partners involved. Not publicly disclosed
Outcomes / Results No published performance metrics or outcomes data. The WPF case study focused on governance process rather than system performance.
Challenges Balancing vendor cost against responsible AI criteria in procurement. Limited staff time to analyse risks and impacts of AI alongside project delivery. Tension between policy aspirations and practical implementation of AI governance tools. Technical details of the ML system remain undisclosed.

How to Cite

DCI AI Hub (2026). 'SUSESO ML-Based Medical Insurance Claims Processing', AI Hub AI Tracker, case CHL-003. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/CHL-003 [Accessed: 1 April 2026].

Change History

Updated 1 Apr 2026, 08:11
by system (system)
Updated 31 Mar 2026, 06:35
by system (system)
Updated 30 Mar 2026, 11:21
by system (system)
Created 30 Mar 2026, 11:18
by v2-import (import)