BEL-001

ONEM Ori Chatbot for Unemployment Benefits Information (Belgium)

Download PDF
Belgium Europe & Central Asia High income Full Production Deployment Confirmed

Office National de l'Emploi / Rijksdienst voor Arbeidsvoorziening (ONEM/RVA), Belgium

At a Glance

What it does Perception and extraction from unstructured inputs — User communication and interaction
Who runs it Office National de l'Emploi / Rijksdienst voor Arbeidsvoorziening (ONEM/RVA), Belgium
Programme ONEM/RVA Unemployment Insurance Information Services
Confidence Confirmed
Deployment Status Full Production Deployment
Key Risks Model-related risks
Key Outcomes Verified outcomes are service-oriented rather than quantitative: Ori provides 24/7 bilingual access to information on unemployment and career break matters, offers guided topic navigation, and routes users to live staff during business hours when automated support is insufficient.
Source Quality 2 sources — Other, Government website / press release

Belgium's National Employment Office, known as ONEM in French and RVA in Dutch, operates the Ori chatbot as part of its unemployment insurance information services. ONEM/RVA is a federal public social security institution responsible for determining eligibility for unemployment benefits, calculating benefit amounts, administering career break and time credit arrangements, supporting jobseeker reintegration, and maintaining nationwide service delivery through a distributed office network. The institution's organisational profile is documented by the European Social Insurance Platform, which describes ONEM/RVA as operating through 30 offices organised into 16 districts and governed through a federal public management structure.

Within that administrative context, Ori functions as a public-facing digital information channel rather than an automated decision-making system. ONEM's own 2025 service page describes the chatbot as a virtual assistant available continuously, twenty-four hours a day and seven days a week, to help users navigate information related to unemployment benefits and career break arrangements. The service is available in both French and Dutch, reflecting Belgium's bilingual administrative environment for this institution. Users can either type questions directly or move through guided prompts and suggested topics. That design indicates a conversational interface intended to reduce friction for citizens who need quick information outside office hours or before escalating to a staff member.

The verified source material supports several operational details that are important for classification. First, Ori is clearly deployed in production on ONEM's live service channels rather than being described as a pilot or internal prototype. Second, the chatbot is positioned as an information and communication tool, not as a mechanism for adjudicating claims, calculating entitlements, or issuing legally binding decisions. Third, ONEM has established a clear escalation path from automated assistance to human support. When the chatbot cannot resolve a user's question, it offers access to live chat with a staff member during business hours. This confirms that human support remains embedded in the service design and that the chatbot is used to triage and answer routine information requests rather than replace civil servants entirely.

The same ONEM source also documents important guardrails around the live escalation workflow. To connect to a live agent, users must provide identifying details through a form, but the staff member assisting through chat cannot make decisions online, cannot share sensitive case-specific information through that channel, and cannot access authenticated transactional services on behalf of the user in that interaction. In practice, this means the chatbot and related live chat function sit on the informational side of service delivery. They help explain procedures, point users toward the right next step, and reduce contact-centre pressure, while sensitive or determinative actions remain separated into other channels with stronger authentication and procedural controls.

The available evidence also supports a cautious interpretation of privacy safeguards. ONEM states that the chatbot itself does not collect personal data during ordinary interactions. The service therefore appears structured around low-risk informational exchanges, with personal data collection limited to the separate live-chat escalation form when a user chooses to contact a staff member. That separation is relevant because it shows how the institution distinguishes between general guidance and case-specific engagement. The service also promotes use of Belgium's secure e-box communication channel for authenticated exchanges with government authorities, reinforcing the division between public information access and protected official correspondence.

This case is therefore retained as a verified example of AI-enabled user communication in social insurance administration. Earlier coding combined Ori with a separate ONEM fraud-detection application described in inaccessible ISSA pages, but those unsupported claims have been removed from the case narrative, programme framing, and coded fields. What remains is the part of the case that can be substantiated from locally available source material: a bilingual, always-available chatbot with human escalation, explicit service limitations, and a clear role in improving citizen access to information about unemployment and career break rules.

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 ONEM/RVA Unemployment Insurance Information Services
Programme Type The type of social protection programme, classified under social assistance, social insurance, or labour market programmes. View in glossary Unemployment 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
Programme Description Belgium's National Employment Office (ONEM/RVA) administers unemployment insurance and related career break services through a national office network. The verified AI-enabled component documented here is the Ori chatbot, which provides bilingual public information and triage support on unemployment and career break topics and can escalate users to live staff during business hours.
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 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
Hybrid Components Automated conversational assistant paired with human live-chat escalation. The exact underlying model architecture is not specified in the verified source material.
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 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 Not documented
Highest Risk Category The most significant structural risk source identified: data, model, operational, governance, or market/sovereignty risks. View in glossary Model-related risks
Risk Assessment Status Whether a formal risk assessment, informal assessment, or independent audit has been conducted for this system. Not assessed

Risk Dimensions

Governance and institutional oversight risks

Impact Dimensions

  • Data minimisation controls
  • Human oversight protocol
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
Beneficiary registries and MISNon-personalSingle source (no linkage)Currently available and usedThe verified service scope is informational rather than transactional. Public-facing content covers unemployment and career break rules and directs users to the appropriate service channels.
Unstructured and text-based contentNon-personalSingle source (no linkage)Currently available and usedOri processes natural language queries from citizens about unemployment regulations and career break provisions. ONEM states that the chatbot does not collect personal data during ordinary interactions. The service operates in French and Dutch and can escalate users to live staff during business hours.

ESIP (n.d.). 'ONEM-RVA', European Social Insurance Platform Members. Available at: https://esip.eu/members/onem-rva (Accessed 26 Mar 2026).

View source Other

ONEM (2025). 'Découvrez comment utiliser notre chatbot Ori et discuter en direct avec un collaborateur', ONEM Actualités, 3 July. Available at: https://www.onem.be/index.php/actualites/2025/07/03/decouvrez-comment-utiliser-notre-chatbot-ori-et-discuter-en-direct-avec-un-collaborateur (Accessed 26 Mar 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 Full Production Deployment
Year Initiated The year the AI system was first initiated or development began. 2025
Scale / Coverage The scale and geographic or population coverage of the deployment. National public-facing service. ONEM/RVA operates 30 offices across 16 districts nationally, and the Ori chatbot is available on the institution's website in French and Dutch for users seeking information on unemployment and career break matters.
Technical Partners External technology vendors, academic partners, or development partners involved. No external vendor or technical partner is identified in the verified source material.
Outcomes / Results Verified outcomes are service-oriented rather than quantitative: Ori provides 24/7 bilingual access to information on unemployment and career break matters, offers guided topic navigation, and routes users to live staff during business hours when automated support is insufficient. The service also channels users toward ONEM's secure communication pathways for authenticated interactions.
Challenges The verified service limitations are also its main constraints: the chatbot cannot make decisions, cannot provide authenticated case handling, and must avoid disclosing sensitive personal information. Accuracy and clarity remain important because unemployment rules are procedurally complex and the service operates in a bilingual environment.

How to Cite

DCI AI Hub (2026). 'ONEM Ori Chatbot for Unemployment Benefits Information (Belgium)', AI Hub AI Tracker, case BEL-001. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/BEL-001 [Accessed: 1 April 2026].

Change History

Updated 31 Mar 2026, 06:35
by system (system)
Created 30 Mar 2026, 08:38
by v2-import (import)