HHS Public Benefits and AI

• Authority: Department of Health and Human Services
• Parent authority: Federal government
• Jurisdiction: United States
• Collections: U.S. regulations, executive orders, and agency policies
• Most recent activity: Enacted
• Most recent activity date: 2024-04-29

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• Unfair discrimination and misrepresentation
• Unequal performance across groups
• Compromise of privacy
• Overreliance and unsafe use
• Loss of human agency and autonomy
• Governance failure
• Lack of capability or robustness
• Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• AI system security vulnerabilities
• False or misleading information
• Fraud, scams, and targeted manipulation
• Increased inequality and decline in employment
• AI pursuing its own goals

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 10-12 subdomains, with strong focus on discrimination and fairness (1.1, 1.3), privacy (2.1), misinformation risks (3.1), human-computer interaction (5.1, 5.2), socioeconomic impacts (6.2, 6.5), and AI system safety (7.1, 7.3, 7.4). Coverage is concentrated in fairness, governance, and system reliability domains relevant to public benefits administration.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• Health Care & Social Assistance
• Public Administration (excl. National Security)

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Professional & Technical Services

All other sectors had no coverage in this document.

Sectors Summary

This document primarily governs the Public Administration sector, specifically state, local, tribal, and territorial government agencies administering HHS-funded public benefits programs. It also has significant coverage of Health Care and Social Assistance due to the nature of many HHS programs. The document indirectly affects Professional and Technical Services (technology vendors) and Scientific Research sectors.

Legislative Status

Soft Law

Legislative Status Summary

This is a non-binding guidance document that provides recommendations to state, local, tribal, and territorial governments. The document explicitly states 'The recommendations laid out in this plan are not mandatory and are general in nature' and uses predominantly voluntary language throughout ('encourages', 'recommends', 'should').

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: Department of Health and Human Services (HHS); Secretary of Health and Human Services
• Summary: The document is authored by HHS in fulfillment of Executive Order 14110 Section 7.2(b) which requires the Secretary of Health and Human Services to publish this plan. HHS is the federal agency proposing these recommendations.

Target

• Entity Types: AI_Governance_Actor, AI_Deployer
• Entities: state, Tribal, local, and/or territorial government entities (STLTs); technology vendors
• Summary: The document explicitly targets state, local, tribal, and territorial governments that administer HHS-funded public benefits programs, as well as their technology vendors. These entities are both governance actors (government agencies) and AI deployers (implementing AI systems in benefits administration).

Enforcer

• Summary: The document does not establish enforcement mechanisms or designate enforcement bodies. It explicitly states that recommendations are not mandatory. While HHS has oversight authority over federally-funded programs, this document does not create new enforceable requirements.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: HHS civil rights and civil liberties offices; HHS program offices
• Summary: HHS indicates it will monitor implementation through engagement with STLTs and may require notification or approval for certain AI uses under existing program requirements. Civil rights offices are referenced for monitoring compliance with nondiscrimination laws.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document comprehensively covers all AI lifecycle stages from planning through ongoing monitoring. It emphasizes design considerations, data quality, model development, testing and validation, deployment practices, and continuous monitoring. The document addresses both AI models and AI systems broadly, with focus on rights-impacting and safety-impacting uses in public benefits administration.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• AI Systems
• Generative AI

Technical Scope Summary

The document explicitly mentions both AI models and AI systems, referring to 'automated or algorithmic systems' and 'AI-enabled automated systems' throughout. It does not specifically mention frontier AI, general purpose AI, foundation models, or compute thresholds. The focus is on AI applications in public benefits administration regardless of model type, with emphasis on rights-impacting and safety-impacting uses.