BRA-001

INSS Helô — AI-Enabled Virtual Assistant for Social Security Services

Download PDF
Brazil Latin America & Caribbean Upper middle income Full Production Deployment Confirmed

Instituto Nacional do Seguro Social (INSS); Dataprev (Empresa de Tecnologia e Informações da Previdência Social)

At a Glance

What it does Perception and extraction from unstructured inputs — User communication and interaction
Who runs it Instituto Nacional do Seguro Social (INSS); Dataprev (Empresa de Tecnologia e Informações da Previdência Social)
Programme Meu INSS (My INSS) Digital Platform — Helô Virtual Assistant
Confidence Confirmed
Deployment Status Full Production Deployment
Key Risks Operational and system integration risks
Key Outcomes Over 100 million interactions processed in less than four years.
Source Quality 7 sources — Government website / press release, Report (multilateral / development partner)

Brazil's National Social Security Institute (Instituto Nacional do Seguro Social, INSS) operates Helô, an AI-enabled virtual assistant deployed on the Meu INSS (My INSS) digital platform, which serves as the primary digital channel for citizens to access social security services across the country. Developed by Dataprev, Brazil's federal social security data processing company, Helô was launched on 7 May 2020 in direct response to the suspension of in-person INSS services on 20 March 2020 during the COVID-19 pandemic. The chatbot was introduced to ensure continuity of service delivery at a time when the INSS was already experiencing reduced personnel capacity, increased demand for services, overloading of its telephone call centre (135 hotline), and the implementation of the 2019 pension reform.

Helô is accessible via both the Meu INSS web portal and the Meu INSS mobile application on Android and iOS devices. Citizens initiate interaction by clicking a chat balloon icon in the lower-right corner of the screen and providing their name and CPF (Cadastro de Pessoas Físicas, the Brazilian individual taxpayer registration number). The system then guides users through structured conversation paths covering a wide range of INSS services and benefits. Core service areas include requesting benefits and services, checking process status, consulting medical examination results, obtaining statements, certificates and declarations, scheduling in-person appointments at INSS agencies, submitting documents, and managing account passwords.

The system was initially implemented using a decision tree architecture with predefined response possibilities for each situation indicated by the user during the interaction. This rule-based approach ensured reliable and consistent responses across the high volume of citizen inquiries. In July 2023, the platform was upgraded to incorporate a more advanced natural language processing (NLP) model, enabling more fluid and personalised conversations that allow citizens to express their queries in natural language rather than navigating through fixed menu options. This evolution from structured decision trees to NLP-based interaction represents a significant technical progression in the system's capabilities. Dataprev has announced plans to further enhance Helô with generative artificial intelligence capabilities.

The scale of Helô's deployment is substantial. On its launch day alone, the system processed 53,595 interactions, and within ten days had handled an additional 191,718 interactions. By May 2021, one year after launch, Helô had surpassed 19 million interactions, averaging approximately 80,000 daily consultations. By October 2020, the system had already reached six million interactions serving approximately three million citizens. By its third anniversary in May 2023, Helô had completed over 66 million interactions, with nine million in 2023 alone across 90 service topics. By 2024, the system had surpassed 100 million interactions in less than four years of operation. Service usage data shows that benefit and service requests constitute the largest share of interactions (approximately 30-35 per cent), followed by statements and certifications (approximately 14 per cent), medical evaluation inquiries (approximately 14 per cent), appointment scheduling (approximately 11 per cent), and password support (approximately 10 per cent).

Helô operates within a hybrid service model that combines automated AI-driven responses with humanised chat support provided by INSS staff. Servers across Brazil manage both Helô's automated operations and the humanised chat component, which requires staff with knowledge of INSS legislation and services. This hybrid approach ensures that more complex citizen inquiries that exceed the chatbot's capabilities are escalated to human operators who can provide fuller responses. The system is being progressively extended to incorporate more sophisticated human-AI collaboration.

Dataprev operates a dedicated AI Centre of Excellence located in João Pessoa, Paraíba, as one of five strategic technology hubs across Brazil. Over 700 developers across these centres support the applications used by citizens daily. In addition to Helô, Dataprev has implemented AI-enabled facial recognition technology within Meu INSS for biometric life verification (prova de vida) of benefit recipients, demonstrating the broader AI integration strategy within Brazil's social security infrastructure.

Helô has received international recognition, winning the 2020 International Social Security Association (ISSA) Best Practices Award for the Americas (ISSA Good Practice gp/198036). The ISSA has highlighted Helô as an example of how social security institutions can leverage chatbot technology to improve citizen service delivery, particularly during crises. The system's accessibility design prioritises simple language free of technical jargon to serve users across all education levels. Planned enhancements include direct document issuance via chat and accessibility improvements such as high contrast mode.

Brazil's broader Meu INSS digital transformation programme has also been recognised by ISSA as a success story in the digital transformation policy of the Brazilian State (ISSA Good Practice gp/198205), providing the institutional and technical foundation upon which Helô operates.

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

Social Protection Functions

Implementation/delivery chain
Outreach/communications/sensitisation primaryProvision of payments/services
SP Pillar (Primary) The social protection branch: social assistance, social insurance, or labour market programmes. Social insurance
Programme Name Meu INSS (My INSS) Digital Platform — Helô Virtual Assistant
Programme Type The type of social protection programme, classified under social assistance, social insurance, or labour market programmes. View in glossary Other
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 Unified digital service channel operated by INSS (Instituto Nacional do Seguro Social) for citizen access to social security services including pension inquiries, benefit requests, medical examination scheduling, document submission, and proof of life. Over 100 million interactions processed in less than four years. Accessible via web portal and mobile application.
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 Hybrid
Lifecycle Stage Current stage in the AI lifecycle, from problem identification through to monitoring, maintenance and decommissioning. View in glossary Monitoring, Maintenance and Decommissioning
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 Developed in-house
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
Hybrid Components Decision tree / rule-based architecture (2020-2023) upgraded to NLP-based natural language processing model (July 2023). Generative AI integration planned.
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
Is Agentic Whether the system autonomously plans and executes multi-step workflows, selecting tools and chaining actions with limited human intervention. View in glossary Partial
Agentic Pipeline Description of the chained workflow steps in the agentic pipeline. Helô autonomously handles the majority of citizen inquiries through structured conversation flows and NLP-based responses covering 90 service topics. Complex queries beyond the chatbot's capability are escalated to human INSS staff via the humanised chat component.
Agentic Autonomy Degree of autonomy: fully autonomous, semi-autonomous (human checkpoints), or supervised (human approval at each step). Semi-autonomous
Override Points Where in the pipeline human review or override is triggered. Human operators handle queries the chatbot cannot address via humanised chat support. INSS staff across Brazil manage escalated interactions requiring knowledge of legislation and services.
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 Low
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 Fully in-house
Highest Risk Category The most significant structural risk source identified: data, model, operational, governance, or market/sovereignty risks. View in glossary Operational and system integration risks
Risk Assessment Status Whether a formal risk assessment, informal assessment, or independent audit has been conducted for this system. Not assessed

Risk Dimensions

Data-related risks
Governance and institutional oversight risks
Operational and system integration risks

Impact Dimensions

Autonomy, human dignity and due process
Equality, non-discrimination, fairness and inclusion
Systemic and societal
  • Human oversight protocol
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
Beneficiary registries and MISSensitiveLinks data across multiple systemsCurrently available and usedINSS beneficiary records and service databases accessed to provide personalised responses about benefit status, process tracking, and appointment scheduling. Linked to INSS administrative systems via Meu INSS platform.
Unstructured and text-based contentPersonalSingle source (no linkage)Currently available and usedCitizen text queries via Meu INSS web portal and mobile application. Users provide name and CPF (individual taxpayer registration number) to initiate service. Chatbot processes natural language inquiries about pensions, benefits, medical examinations, certifications, and appointments.

Dataprev (2020) 'Chatbot Helô, do Meu INSS, já fez seis milhões de atendimentos', dataprev.gov.br.

View source Government website / press release

Dataprev (2020) 'Conheça Helô, a atendente virtual que a Dataprev desenvolveu para o INSS', dataprev.gov.br.

View source Government website / press release

Dataprev (2021) 'Helô completa 1 ano com mais de 19 milhões de atendimentos no Meu INSS', portal3.dataprev.gov.br.

View source Government website / press release

Dataprev (n.d.) 'Dataprev investe em IA, que está em serviços como a Helô, do Meu INSS', portal3.dataprev.gov.br.

View source Government website / press release

INSS (2023) 'Helô, assistente virtual do INSS completa três anos', gov.br/inss.

View source Government website / press release

INSS (2024) 'Helô comemora 100 milhões de acessos em menos de 4 anos', gov.br/inss.

View source Government website / press release

ISSA (n.d.) 'Helô, the National Social Security Institute's virtual assistant', issa.int, Good Practice gp/198036.

View source Report (multilateral / development partner)
Deployment Status How far the system has progressed into real-world operational use, from concept/exploration through to scaled and institutionalised. View in glossary Full Production Deployment
Year Initiated The year the AI system was first initiated or development began. 2020
Scale / Coverage The scale and geographic or population coverage of the deployment. National coverage across Brazil (population ~214 million). Over 100 million interactions in less than four years. Approximately 80,000 daily consultations (2021 average). 66 million cumulative interactions by May 2023. Nine million interactions in 2023 alone.
Funding Source The source(s) of funding for the AI system development and deployment. Government of Brazil (Dataprev / INSS budget)
Technical Partners External technology vendors, academic partners, or development partners involved. Dataprev (Empresa de Tecnologia e Informações da Previdência Social) — AI Centre of Excellence, João Pessoa, Paraíba
Outcomes / Results Over 100 million interactions processed in less than four years. 53,595 interactions on launch day. 19 million interactions in first year. 66 million cumulative interactions by May 2023. Approximately 80,000 daily consultations. Winner of 2020 ISSA Best Practices Award for the Americas. Enabled continuity of social security services during COVID-19 pandemic when in-person services were suspended.
Challenges Initial decision tree architecture limited to predefined response paths, requiring upgrade to NLP in July 2023. Hybrid model requires INSS staff with specialised legislative knowledge to handle escalated queries. Generative AI integration still in planning phase. Specific NLP model architecture and training data details are not publicly documented. Reports have emerged of broader AI-enabled features in the Meu INSS app wrongly rejecting benefit claims, though this relates to separate AI components rather than the Helô chatbot specifically.

How to Cite

DCI AI Hub (2026). 'INSS Helô — AI-Enabled Virtual Assistant for Social Security Services', AI Hub AI Tracker, case BRA-001. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/BRA-001 [Accessed: 1 April 2026].

Change History

Updated 1 Apr 2026, 08:11
by system (system)
Created 30 Mar 2026, 08:38
by v2-import (import)