Social Security Administration (SSA) – Quick Disability Determinations (QDD) Predictive Model
Overview
The United States Social Security Administration (SSA) operates a system known as Quick Disability Determinations (QDD), which uses a computer-based predictive model to expedite the processing of some disability claims within the Social Security Disability Insurance (SSDI) programme. The QDD process screens initial disability applications to identify cases where a favourable disability determination appears highly likely and where medical evidence is readily available. When a claimant completes a disability application, SSA creates an electronic file containing their information and medical history. The QDD predictive model then scans the file for keywords and phrases that indicate the claimant is likely to be found disabled, including medical conditions that might meet one of SSA's published medical listings. If the programme identifies these keywords, it checks to ensure the application has all the necessary documentation before flagging the case for expedited processing.
QDD is one of SSA's longstanding fast-track disability processes, alongside Compassionate Allowances, which together represent what SSA describes as two of its greatest successes in recent years, allowing the agency to approve some cases in a matter of days instead of months. Official SSA materials describe QDD as a predictive screening tool that has been used nationally since February 2008 and is continually refined to reflect the characteristics of the recent applicant population. SSA is described as having actively employed artificial intelligence for over 20 years, with 14 current AI use cases as of 2024, making QDD part of a broader institutional commitment to data-driven decision support.
The retained sources support the conclusion that QDD functions as a screening and workload-prioritisation model rather than an automated benefits decision-maker. Applications flagged by the model are sent to a special group of QDD claims examiners, who review the file and make the actual determination. A disability examiner should start looking at a flagged file a day or two after receiving it. If the medical records are complete and the QDD unit agrees with the claimant's alleged onset date, the case can be approved in less than a month. However, if the record does not contain enough medical evidence or the date of disability onset is harder to determine, the QDD examiner may remove the application from the fast track and return it to standard processing. This does not result in a denial but means the claim will be processed through SSA's usual channels.
SSA has maintained formal operating instructions for staff handling QDD cases, including specific procedures for field offices and Disability Determination Services, and has continued to treat the model as an active operational tool rather than a one-off pilot. At the same time, SSA does not publicly disclose much detail about the current model logic, feature set, fairness testing, or performance drift management. The case is therefore best understood as a real and long-running high-consequence screening deployment with limited public technical transparency.
Publicly available secondary sources report materially faster processing times for QDD cases than for standard disability claims. According to the most recent SSA statistics available as of December 2022, QDD cases were taking an average of 27 days, with a median of only 13 days. SSA's chief artificial intelligence officer reported a 157 percent year-over-year jump in process rates for the disability screening tool and noted plans for expanded use across SSA teams. Those figures are directionally consistent with SSA's official framing that QDD accelerates clear-cut claims, but the exact contemporary coverage and selection-rate metrics are not richly disclosed in official primary materials.
Classification
AI Capabilities
Use Cases
Social Protection Functions
| SP Pillar (Primary) | Social insurance |
Programme Details
| Programme Name | Social Security Administration (SSA) – Quick Disability Determinations (QDD) Predictive Model |
| Programme Type | Old age, survivors and disability pensions |
| System Level | Implementation/delivery chain |
The Social Security Disability Insurance (SSDI) programme administered by the U.S. Social Security Administration, which provides benefits to individuals who are unable to work due to a qualifying disability. QDD is a fast-track process within this programme that uses a predictive model to identify and expedite clear-cut disability claims.
Implementation Details
| Implementation Type | Classical ML |
| Lifecycle Stage | Monitoring, Maintenance and Decommissioning |
| Model Provenance | Developed in-house |
| Compute Environment | Not documented |
| Sovereignty Quadrant | Not assessed |
| Data Residency | Not documented |
| Cross-Border Transfer | Not documented |
Risk & Oversight
| Decision Criticality | High |
| Human Oversight | HITL |
| Development Process | Fully in-house |
| Highest Risk Category | Not assessed |
| Risk Assessment Status | Not assessed |
Risk Dimensions
Data-related risks
Governance and institutional oversight risks
Model-related risks
Operational and system integration risks
Impact Dimensions
Autonomy, human dignity and due process
Equality, non-discrimination, fairness and inclusion
Safeguards
Deployment & Outcomes
| Deployment Status | Scaled & Institutionalised |
| Year Initiated | 2006 |
| Scale / Coverage | National — used across all SSA disability claims nationally since February 2008 |
| Funding Source | Unknown |
| Technical Partners | Custom-developed and operated internally by SSA; no external commercial vendor is identified in primary documentation. |
Outcomes / Results
Official SSA materials and secondary reporting indicate that QDD is used nationally to expedite a subset of likely-favourable disability claims and that these cases are processed materially faster than standard disability claims. Public reporting has cited median handling times measured in days rather than months, but SSA's public primary materials provide only limited detail on current exact performance and coverage rates.
Sources
- SRC-004-USA-007 DisabilitySecrets (n.d.) 'What Is the Quick Disability Determination (QDD) Process?'. Available at: https://www.disabilitysecrets.com/resources/disability/quick-disability-determination-process-social- (Accessed: 23 March 2026).
https://www.disabilitysecrets.com/resources/disability/quick-disability-determination-process-social- - SRC-002-USA-007 Glaze, K., Ho, D.E., Ray, G.K. and Tsang, C. (n.d.) 'Artificial Intelligence for Adjudication: The Social Security Administration and AI Governance', Stanford University Digital Health Observatory. Available at: https://dho.stanford.edu/wp-content/uploads/SSA.pdf (Accessed: 23 March 2026).
https://dho.stanford.edu/wp-content/uploads/SSA.pdf - SRC-003-USA-007 Bracken, M. (2024) 'Data tracking, AI and modernization: How SSA has ditched its paper-based past', FedScoop, 5 July. Available at: https://fedscoop.com/social-security-administration-securitystat-ai-modernization/ (Accessed: 23 March 2026).
https://fedscoop.com/social-security-administration-securitystat-ai-modernization/ - SRC-001-USA-007 U.S. Social Security Administration (2019) Quick Disability Determinations (QDD). Available at: https://www.ssa.gov/disabilityresearch/qdd.htm (Accessed: 23 March 2026).
https://www.ssa.gov/disabilityresearch/qdd.htm
How to Cite
DCI AI Hub (2026). 'Social Security Administration (SSA) – Quick Disability Determinations (QDD) Predictive Model', AI Hub AI Tracker, case USA-007. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/USA-007