USA-004

Artificial Intelligence Adjudicator Assistance (AIAA) - UI Adjudication Prototype

Download PDF
United States North America High income Pilot / Controlled Trial Phase Confirmed

U.S. Department of Labor (DOL), Employment and Training Administration (ETA), Office of UI Modernization

At a Glance

What it does Classification — Decision support for eligibility and benefits
Who runs it U.S. Department of Labor (DOL), Employment and Training Administration (ETA), Office of UI Modernization
Programme Unemployment Insurance (UI) Adjudication - AIAA Prototype
Confidence Confirmed
Deployment Status Pilot / Controlled Trial Phase
Key Risks Governance and institutional oversight risks
Key Outcomes At prototype stage.
Source Quality 2 sources — News article / media, Government website / press release

Artificial Intelligence Adjudicator Assistance (AIAA) is a U.S. Department of Labor research and prototyping initiative exploring whether AI tools could help unemployment-insurance adjudicators sort cases and focus effort on claims that require more fact-finding. The retained sources clearly support that this is a real federal initiative, developed with Stanford RegLab and the Colorado Department of Labor and Employment, but they also make clear that it is not a production decision system. For production-quality writing, the case should therefore remain firmly framed as a prototype and learning exercise.

The initiative emerged from the operational stress that unemployment-insurance systems experienced during the pandemic, when states faced very large spikes in claims and struggled with staffing and outdated technology. During the onset of the COVID-19 pandemic, initial unemployment-insurance claims spiked by 3,000 percent in a matter of weeks, rising from 220,000 per week to more than 6 million and staying above 1 million per week for a year. Responding to this sudden and dramatic increase was extremely difficult for state UI programmes, with limited staffing, constrained resources, and old technology identified as the biggest challenges. The White House Executive Order 14110 on Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, issued on 30 October 2023, further underscored the priority of responsible AI use for federal agencies and provided additional impetus for DOL's initiative.

According to DOL, AIAA is intended to explore whether AI can help adjudicators distinguish between claims requiring extensive fact-finding and those that may be simpler to process, and whether it can assist with extracting or routing relevant information from historical case materials. In UI, adjudication is the process of reviewing claims to determine if they meet eligibility criteria according to state and federal regulations. Adjudicators review applications but often need additional information to determine eligibility, and a significant part of their duties involves conducting fact-finding efforts such as interviewing claimants and employers and submitting requests for additional information. Some eligibility issues require significant fact-finding while others require minimal or no fact-finding. Being able to separate claims based on how much fact-finding they require could bring significant efficiencies. By streamlining the adjudication process, AI could ultimately prevent unnecessary back-and-forth between a claimant and a state UI agency, which stresses an already strained system and can cause delays in eligibility determinations or benefit payments, sometimes leaving claimants waiting for weeks or months.

The strongest official evidence shows that the prototype is being built and tested using historical Colorado unemployment-insurance claims in a locked environment. DOL and RegLab describe a process in which senior claims examiners review and re-adjudicate historical claims to help generate higher-quality training and evaluation material. Colorado's Department of Labor and Employment is providing historical claims data and working with DOL's research partners at Stanford University to test how AI could have potentially assisted with that universe of past data, comparing the model's results to human expertise past and present. Andrew Stettner, director of DOL's Office of UI Modernization, stated that the focus is on how technology can assist the staff that work on UI programmes to do the work more accurately and efficiently, rather than replacing human intelligence. DOL has communicated that it plans to document the work to help states learn about the process of developing an AI model, including the things that an AI model does well and the things that it does not do well. In addition to the UI adjudication prototype, DOL and RegLab are also collaborating on a trustworthy AI guide and a separate pilot of tools for adjudicating workers' compensation claims.

This means the case is notable not because of scale or current operational impact, but because it is an unusually well documented example of a federal agency experimenting cautiously with adjudication support. The initiative is explicitly positioned as using the current period of low unemployment to prepare the system for the next surge. The retained sources do not justify stronger claims about model type, production readiness, or measured performance. They do support the conclusion that DOL is treating the work as a bounded prototype with human adjudicators remaining fully responsible for eligibility determinations. The case therefore remains valid, but only as an early-stage adjudication-support prototype rather than a mature deployment.

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
SP Pillar (Secondary) The social protection branch: social assistance, social insurance, or labour market programmes. Labour market programmes
Programme Name Unemployment Insurance (UI) Adjudication - AIAA Prototype
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
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 U.S. Unemployment Insurance programme administered by state workforce agencies under federal oversight by the Department of Labor, Employment and Training Administration. The AIAA prototype targets the adjudication function within this programme, specifically the triage and fact-finding components of eligibility determination for UI claims.
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 Model Selection and Training
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 Mix of in-house and third-party
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. Informal assessment

Impact Dimensions

Equality, non-discrimination, fairness and inclusion
  • Exit/rollback plan
  • Human oversight protocol
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
Beneficiary registries and MISSensitiveSingle source (no linkage)Currently available and usedHistorical UI claims data from Colorado only; used in a locked environment for prototype training and testing; includes adjudication decisions and claim-document metadata
Unstructured and text-based contentSensitiveSingle source (no linkage)Currently available and usedClaim documents and fact-finding records from historical Colorado UI cases; senior claims examiners re-adjudicated cases to create high-quality labelled training data

Nextgov/FCW (2024) 'Labor Department experiments with AI in unemployment systems', Nextgov/FCW, 20 February. Available at: https://www.nextgov.com/digital-government/2024/02/labor-department-experiments-ai-unemployment-systems/394179/ (Accessed: 24 March 2026).

View source News article / media

U.S. Department of Labor, Employment and Training Administration (2024) 'Introducing Artificial Intelligence Adjudicator Assistance (AIAA): A Research Initiative Exploring Ways to Streamline Work for Adjudicators'. Washington, DC: U.S. Department of Labor. Available at: https://www.dol.gov/agencies/eta/ui-modernization/aiaa (Accessed: 24 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 Pilot / Controlled Trial Phase
Year Initiated The year the AI system was first initiated or development began. 2024
Scale / Coverage The scale and geographic or population coverage of the deployment. Prototype using historical claims data from the state of Colorado only; not deployed at scale
Funding Source The source(s) of funding for the AI system development and deployment. U.S. Department of Labor federal funding
Technical Partners External technology vendors, academic partners, or development partners involved. Stanford University Regulation, Evaluation, and Governance Lab (RegLab); Colorado Department of Labor and Employment (CDLE). No commercial vendor identified.
Outcomes / Results At prototype stage. DOL states the goal is to document what the model does well and what it does not do well. No publicly disclosed quantified performance outcomes yet. The initiative is explicitly framed as a learning exercise to help states understand the process of developing an AI model in the UI context.
Challenges COVID-19 pandemic exposed severe capacity constraints in state UI systems (claims spiked 3,000% in weeks); states have limited staffing, constrained resources, and outdated technology; adjudication backlogs cause delays in eligibility determinations and benefit payments; need to balance AI innovation with trustworthy and responsible deployment practices.

How to Cite

DCI AI Hub (2026). 'Artificial Intelligence Adjudicator Assistance (AIAA) - UI Adjudication Prototype', AI Hub AI Tracker, case USA-004. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/USA-004 [Accessed: 1 April 2026].

Change History

Updated 31 Mar 2026, 06:35
by system (system)
Created 30 Mar 2026, 08:42
by v2-import (import)