Policy & Practice | Fall 2026

The Magazine of the American Public Human Services Association Fall 2026

Transforming Systems Through Technology and Innovation

Prompt:

TODAY’S EXPERTISE FOR TOMORROW’S SOLUTIONS

www.aphsa.org

contents

Vol. 84, No. 3 Fall 2026

departments

features

16

20

4 From Our Partners

Permission, Not Paperwork: Rewiring Eligibility Verification for the People Human Services Serves

6 From Our Partners

Leveraging Verification to Create the Future of Benefits Eligibility: Why Purpose-Built Verification Is the Foundation for Modernizing Human Services Administration

Strengthening SNAP Performance Today Building the Foundation for Modernization

The Human Side of Human Services Modernization How Human Services Leaders Can Strengthen Adoption, Trust, and Performance by Advancing Workforce Enablement and Agentic AI Together

8 Partnering for Impact

Building the Bridge: How Community Pathways and Modern Technology Can Advance FFPSA Goals

10 Partnering for Impact

Breaking Down Silos: How Data SAIL Is Building More Connected, Customer-Centered Services

24

28

12 AI in Focus

Laying the Groundwork for AI Involvement: Why Data Governance Is the Ultimate Prerequisite

14 AI in Focus

A New Path Forward for Eligibility Modernization: How AI-Assisted Development, Applied with Deep Expertise, Can Make Modular Modernization a Reality

From Modular Build to Enterprise Stewardship Operating New Mexico’s Modern Health and Human Services Ecosystem

Fighting Fraud Through System Design What Wyoming’s Experience Reveals About the Future of Child Care Program Integrity

36 The Human in Human Services Back to the Future: Bridging the Technology Gap for Returning Citizens

38 Technology Speaks

Modernizing in Place: Building an Intelligence Layer Around Human Services Systems

32

40 Technology Speaks

Unlocking the Unstructured: How Narrative Analytics Is Completing the Promise of Data-Driven Child Welfare

42 Association News

Honoring the 2026 EMWB Impact Award Winners

Getting CCWIS Across the Finish Line What Successful States Did Right

1

Fall 2026 Policy & Practice

APHSA Executive Governing Board

Strategic Industry Partners DIAMOND

Elected Director Christine Norbut Beyer, Executive in Residence, Think Of Us Elected Director Charlie Brereton , Director, Montana Department of Public Health and

Chair Grace Hou, Deputy Governor for Health and Human Services, State of Illinois Vice Chair Kristi Putnam, Executive Policy Advisor Treasurer Sherron Rogers , Vice President, Finance, Children’s Hospital of Philadelphia Affinity Group Representative Karen Barber, General Counsel, Vermont Department of Mental Health Leadership Council Representative Jeffrey Cartmell, Executive Director, Oklahoma Human Services Local Council Representative Amy Swift, Director, Spotsylvania County (VA) Department of Social Services Elected Director Rodney Adams, Principal, R Adams & Associates Elected Director Derrik Anderson, Executive Director, Race Matters for Juvenile Justice

Human Services Elected Director Vannessa L. Dorantes, Managing Director, Casey Family Programs Elected Director Tracy Gruber , Executive Director, Utah Department of Health and Human Services Elected Director Eboni Washington , Director of Government and Community Relations, Action for Child Protection Immediate Past Chair Dannette R. Smith, Former Commissioner, Colorado Behavioral Health Administration and Former CEO, Nebraska Department of Health and Human Services President & CEO Reggie Bicha, President & CEO, APHSA

PLATINUM

SILVER

2

Policy & Practice Fall 2026

Policy & Practice™ (ISSN 1942-6828) is published four times a year by the American Public Human Services Association, 901 North Glebe Road, Suite 210, Arlington, VA 22203. For subscription information, contact APHSA at (202) 682-0100 or visit the website at www.aphsa.org. Copyright © 2026. All rights reserved. This magazine may not be reproduced in whole or in part without written permission from the publisher. The viewpoints expressed in contributors’ materials are the authors’ own and do not necessarily reflect the policies or views of APHSA. Postmaster: Send address changes to Policy & Practice 901 North Glebe Road, Suite 210, Arlington, VA 22203

Advertising Anna Nogueira anogueira@aphsa.org Design & Production Chris Campbell

President & CEO Reggie Bicha Communications Director Jessica Garon jgaron@aphsa.org

Editor Amy Plotnick

2026 Advertising Calendar

Issue

Ad Deadline October 17

Issue Theme

Winter

Building Stronger Communities Together

Size and Placement Two-page center spread: Back Cover (Cover 4): Inside Back Cover (Cover 3): Inside Front Cover (Cover 2):

Rate

Discount for 4 Consecutive Issues

$8,000 $6,000 $5,400 $5,400 $3,000 $1,200 $1,050

$7,200/issue $5,400/issue $4,860/issue $4,860/issue $2,700/issue $1,080/issue $945/issue

Full page:

Half page, horizontal: One-third page, vertical:

3

Fall 2026 Policy & Practice

By Justin Stolzenberg from our partners

Permission, Not Paperwork: Rewiring Eligibility Verification for the People Human Services Serves

T he job of an eligibility caseworker has not changed much in recent years. A household applies. The case worker has 30 days, or 7 for expedited cases, to verify income, employment, household composition and, some times, assets. Most states have moved past paper-only methods and lean on the dominant employer database. But it isn’t comprehensive. Part-time work, gig income, and self-employment fall outside its scope, and the data it does return can be weeks out of date. Out come the pay stubs, the employer verification forms, the bank state ments, or requests for them. The result entails high verification costs, deci sions made on incomplete information, and a workload that still falls back on manual paperwork. This process supports tens of millions of households and hundreds of billions in benefits. It was built for a workforce that no longer exists. More than 60 million Americans now do some form of independent work each year, according to MBO Partners, piecing together income from W-2 jobs, gig platforms, side busi nesses, and seasonal work. Household finances have multiplied too, spread across banks, neobanks, prepaid cards, and peer-to-peer wallets. The eligibility systems behind the Supplemental Nutrition Assistance Program (SNAP), Medicaid, Temporary Assistance for Needy Families (TANF), child care subsidies, the Low-Income A WorkforceThat Has Changed Faster Than Our Systems

Home Energy Assistance Program (LIHEAP), Supplemental Security Income (SSI), and dozens of state programs were not designed for house holds juggling multiple employers, irregular hours, gig platforms, and fragmented accounts. The mismatch shows up everywhere — in procedural denials, in tens of billions of dollars in improper payments flagged by the U.S. Government Accountability Office each year and, most painfully, in the Medicaid unwinding, where roughly two-thirds of disenrollments were pro cedural rather than substantive. Under H.R. 1, states with elevated SNAP payment error rates face direct cost-share exposure beginning in fiscal year 2028. Modernizing verifica tion has shifted from an operational priority to a fiscal one. From Inquiry to Permission For decades, eligibility verifica tion has been inquiry based. Agencies

ask, then piece together answers from employer outreach, third-party databases, and applicant-submitted documents. The applicant’s role is to produce paperwork. Consent-based verification (CBV) reverses that. The applicant grants explicit, time-bound consent for an agency to receive their income, employment, and asset data directly from the source. The data are current, structured, and certified by the applicant. Data quality improves when agencies see the full structure of someone’s earnings, not a static snapshot. Coverage improves when modern direct-source connectivity reaches roughly 96 percent of the U.S. work force, including 25+ gig platforms and thousands of financial institutions. The legacy employment database returns a match for only about a third of inqui ries. And the applicant moves back to the center. A common objection is that

Illustration by Chris Campbell

4

Policy & Practice Fall 2026

at the center of every determination. Queues are already over capacity, and policy changes ahead will only add to the load. What CBV removes is the friction around the determination. The pending cases waiting on a missing pay stub, the unanswered employer calls, and the document scramble before a federal deadline. Removing that friction does not shrink the work force. It returns caseworker time to the work only they can do, supporting the families in front of them. In modeled scenarios, that shift returns roughly 500,000 caseworker hours to higher-value work and produces $11 million in direct verifica tion savings annually, with error rate reduction adding as much as $115 million more. This is operating today in state benefits programs. The Opportunity in Front of Us State leaders do not have to start from scratch when evaluating this category. 17A’s Consent-Based Verification: A Buying Guide for State Medicaid and SNAP Programs 1 lays out the questions worth pressing on. Which payroll providers, gig platforms,

and financial institutions does the vendor cover today? What share of applicants offered CBV consent and complete it? Is client data auto populated into existing state systems, or does staff still re-key it? And what paper trail does the tool leave behind for audits and case notes? The directors I talk to are not looking for transformation as a concept. They want five minutes back in their day, a determination that holds up at audit, and a verification process that does not put their state on the wrong side of the new federal cost-share rules. That is what is now available and proven. 2 The question is no longer whether this category deserves serious consider ation. It is how agencies evaluate it responsibly, implement it thoughtfully, and use it to reduce burden without compromising trust.

consent flows add friction. The objec tion has it backwards. The current model lets states purchase access to applicants’ payroll data on the strength of a general authorization buried in benefits paperwork; the applicant rarely sees what was shared, with whom, or for how long. CBV replaces that with explicit, time-bound, revocable consent. WhatThis Looks Like at Scale For a hypothetical state running 3 million Medicaid and SNAP verifica tions a year, legacy methods match about 31 percent of inquiries. The remaining 2.1 million cases fall to manual review, consuming roughly 700,000 caseworker hours and $16 million annually. A waterfall that starts with CBV looks different. About 55 percent of applicants verify directly from the source in seconds. Another 20 to 25 percent resolve through optical character recognition (OCR)–based document upload. Only the residual cases fall to manual casework. None of this replaces eligibility caseworkers. Their judgment, empathy, and program expertise sit

Justin Stolzenberg is the Vice President at Argyle.

Reference Notes 1. https://drive.google.com/file/d/10YB GbyZN8A3dSnn1rFlksZsebgtQCq2T/ view?usp=sharing 2. https://www.argyle.com/blog/states-are moving-to-cbv-choose-a-partner-thats-proven

5

Fall 2026 Policy & Practice

By Melissa Wolf from our partners

Leveraging Verification to Create the Future of Benefits Eligibility: Why Purpose-Built Verification Is the Foundation for Modernizing Human Services Administration

F or 22 years, I sat on the “other side of the table.” As a leader of human services programs like Medicaid and the Supplemental Nutrition Assistance Program (SNAP), I lived the reality that most agency directors face today: the constant pressure of shrinking timelines, the weight of massive case loads, and the persistent, nagging frustration of seeing applicants in moments of genuine need stalled by paperwork that was meant to help, not hinder, them. During my tenure, I reviewed countless technology solu tions that promised to “modernize” service delivery and lessen the pres sures that my team and I faced, but the promises often fell short. The reason is that many of these solutions were developed for use in other industries and then adapted for government. With stakes so high, as human services agencies across the country work to implement the expanded eligi bility requirements mandated by H.R. 1, improve accuracy, and absorb rising workloads, the verification technology they choose, and the genesis of the solution and what it was built to do, and for whom, matters more than ever. Why “Purpose-Built” Verification Is Critical for Human Services Agencies A purpose-built verification solution is one engineered, from the ground up, to address the unique regulatory and administrative complexities of human services, rather than simply adapting verification tools originally built to address other industries for

government use. A purpose-built approach ensures that the tools themselves are designed to reflect the specific realities of public benefit programs and drive outcomes that matter most, including accuracy, program integrity, and timeliness to deliver. By creating technology in deep collaboration with agency leaders, program experts, and applicants with lived experience, purpose-built solu tions speak the language of public services. They are designed to intervene at the most critical points of the eligi bility process, ensuring that agencies can prioritize program integrity and accuracy without creating unnecessary friction for those they serve. A purpose-built approach acknowl edges three key “customer groups” who must be served:

1. The Agency: The human services agency, which needs to maintain program integrity and meet federal reporting standards while regaining administrative capacity. 2.The Applicant: The applicant, who needs to submit income, hours, and employment to prove eligibility without being subjected to burden some “document chasing” that feels like a barrier. 3. The Taxpayer: The broader public, which expects and deserves govern ment efficiency and program integrity. Questions to Ask When Selecting a Verification Partner Choosing a verification partner is a strategic decision that goes beyond pro curement; it is a fundamental choice about how your agency will operate.

Illustration by Chris Campbell

6

Policy & Practice Fall 2026

We are in a unique moment in history. The mandated, expanded H.R. 1 requirements are not only a shift in policy, but a stress test for the entire public benefit infrastructure.

Too often, agencies fall into the trap of evaluating technology based on a “features list” without understanding the impact on workflow and how it dictates eligibility determinations will be made when implemented. If you are an agency leader currently evaluating potential verification solu tions, I encourage you to ask the hard questions that reveal whether a product is truly purpose-built and designed for the mission of public services: n Does the solution provide the com prehensive data reporting you need? Is anything missing? n What is the realistic timeline for implementation? n Can the product be customized or will it dictate workflow and how your agency determines benefit eligibility? n Can the solution support multiple programs (e.g., SNAP, Medicaid, Temporary Assistance for Needy Families) simultaneously? n How does the solution help you mitigate known areas of friction in the eligibility determination process? n How does the solution help eliminate opportunities for errors to occur? n How does the solution help regain staff capacity?

agencies need more than just any tech nology. They need a solution built for their mission. Effective eligibility modernization requires us to demand better from the technology we use. By choosing verifi cation solutions designed specifically for human services by former human services leaders who have carried the weight of this work and are informed by benefit recipients with lived expe rience, agencies aren’t just verifying data. They are serving their citizens with more speed, more accuracy, and, ultimately, more compassion. And that, ultimately, is how purpose-built technology will shape the future of benefits eligibility and modernize human services administration. Melissa Wolf is the Director, Government Strategy & Operations, at SteadyIQ.

n What is built into the solution to drive outcomes that matter, such as improved accuracy, decreased benefit delivery time, and reduced administrative burden? n What is the ease of use for both agency staff and applicants? n What level of training support does the vendor provide? Do trainers understand what is important to agencies? n How are product enhancements handled, particularly as new policy updates occur? Meeting the Moment With Purpose We are in a unique moment in history. The mandated, expanded H.R. 1 requirements are not only a shift in policy, but a stress test for the entire public benefit infrastructure. To pass this test, human services

7

Fall 2026 Policy & Practice

partnering for impact

By Alana Barr, Jennifer O’Brien, and Yolanda Green-Rogers

Building the Bridge: How Community Pathways and Modern Technology Can Advance FFPSA Goals

T he Family First Prevention Services Act (FFPSA), enacted in 2018, marked one of the most signifi cant shifts in child welfare policy in decades. Rather than focusing on foster care placement after a crisis occurs, FFPSA encourages states and jurisdic tions to invest in prevention services that strengthen families and prevent out-of-home care. This transformation represents more than a policy adjust ment—it requires an entirely new operational model for child welfare agencies and community partners. At the center of this evolution is the concept of a community pathway. 1 While the vision is compelling, implementation presents substantial technological and operational chal lenges. To succeed, agencies must move beyond fragmented referral processes, disconnected data systems, and manual coordination methods. Technological solutions capable of supporting real-time collaboration, closed-loop referrals, provider engage ment, and fidelity or outcomes tracking are becoming essential infrastructure for FFPSA implementation. This is where innovative closed- loop service referral management systems, such as those developed by Binti, Inc., are poised to play a trans formative role. FFPSA and the Shift Toward Upstream Prevention FFPSA fundamentally changed how states can use federal child welfare funding. By allowing Title IV-E reimbursement for approved evidence based practices (EBPs), including mental health counseling, substance

for operationalizing FFPSA goals and shifting prevention further upstream. At their core, community pathways create networks that offer families services and case management outside of child welfare. Across the country, states and jurisdictions are exploring different approaches for families to access voluntary prevention services through any number of doors, such as through school counseling, health care professionals, housing organizations, or nonprofit organizations, without having to interact with the child welfare agency. Community pathway implementation requires agencies to collect a minimal amount of data to preserve family privacy and ensure access to stabilizing services.

use services, and in-home parent skill based programs, and kinship navigator programs, the law incentivizes agencies to intervene earlier and more effec tively. Historically, many child welfare systems operated reactively. Families often encountered formal systems only after crises escalated to abuse or neglect investigations. FFPSA seeks to reverse that dynamic by encouraging cross-system collaboration among various entities, such as child welfare agencies, schools, health care orga nizations, courts, behavioral health providers, and community-based organizations. However, prevention oriented systems require a very different operational architecture than traditional placement-focused systems. Community pathways are emerging as practical frameworks

See FFPSA Goals on page 44

Illustration by Chris Campbell

8

Policy & Practice Fall 2026

A TRAINING HUB A TRAINING HUB FOR RESOURCES, FOR RESOURCES, INNOVATION, AND INNOVATION, AND VIRTUAL VIRTUAL EXCHANGE EXCHANGE

learn to thrive

INTERACTIVE, SELF-PACED E-LEARNING COURSES resources built for human services professionals online learning communities

DOWNLOAD OUR MOBILE APP

Android

apple

thrive.matrixlms.com

9

Fall 2026 Policy & Practice

partnering for impact

By Nicole Acosta

Breaking Down Silos: How Data SAIL Is Building More Connected, Customer-Centered Services

F or individuals and families navi gating public benefits and services, getting the support they need can mean interacting with multiple agencies, programs, and systems. Too often, those systems do not communicate with one another, creating additional work for both the people seeking services and the staff supporting them. The Data & Impact State Action & Innovation Lab (Data SAIL) is helping states change that. Data SAIL is the implementation phase of the Aligned Customer Centered Ecosystem of Services & Supports (ACCESS) initiative, led by the American Public Human Services Association (APHSA) and the National Association of State Workforce Agencies (NASWA). ACCESS advances a vision in which health, human services, and workforce systems work together to create more coordinated, customer-centered experiences. Through Data SAIL, states are putting that vision into practice by developing data-enabled solutions that improve program operations and outcomes for individuals and families. Following a rigorous selection process, Kentucky, Maryland, and Nebraska were selected to partici pate in Data SAIL. While each state is tackling a different challenge, all three projects focus on breaking down barriers between programs, data, and systems for improved customer experi ences and community outcomes Bringing Together the Right Expertise Data SAIL convenes a robust eco system of collaborators throughout

In early 2026, APHSA and its technical assistance (TA) partners convened the three cross-sector state teams in Austin, TX to jumpstart team collaboration and cohort learning. CIC members led workshops on inte grating customer perspectives into projects, and co-developed project logic models with state teams. State teams continue to engage in cohort learning with APHSA and its TA partners. Kentucky: Connecting Referrals to Outcomes Through Data SAIL, the Kentucky Cabinet for Health and Family Services (CHFS) and Education and Labor Cabinet (ELC) are working together to understand how refer rals addressing social determinants

implementation. APHSA provides human services expertise and facilitates cross-state learning, while NASWA brings workforce expertise, state networks, and knowledge of employment and training programs. U.S. Digital Response (USDR) provides technical expertise to help states address complex digital, data, and operational challenges. Coleridge supports state agencies in strengthening their ability to use data effectively. The ACCESS Community Impact Council (CIC), made up of eight individuals with lived expertise accessing human services, brings an essential customer perspective to the work. CIC members engage with state teams through cohort calls and in person events to help ensure solutions reflect the experiences of people who navigate these systems.

10

Policy & Practice Fall 2026

of health connect to education and employment outcomes. The project centers on kynect resources, Kentucky’s statewide resource directory and closed-loop referral platform. Residents can find community resources, complete a social determinants of health (SDoH) assessment, and track referrals, while community partners can coordinate services and document outcomes. Kentucky also partners with United Way 2-1-1, whose referral special ists help ensure referrals do not go unanswered. Through quantitative analysis and qualitative user experience evalu ation, Kentucky is examining what is working, identifying gaps, and exploring how services can better respond to residents’ needs. The project is also strengthening collabo ration across health, human services, education, and workforce systems. Maryland: Improving Data Matching for SUN Bucks Maryland is developing an AI-driven “MatchBot” solution to improve data matching for the state’s Summer EBT, or SUN Bucks, program, which provides grocery benefits to eligible families with school-aged children during the summer. When records cannot be matched automatically, case managers may need to manually compare informa tion across systems. Incomplete or inconsistent customer information can make this process time-consuming and less accurate. Maryland’s MatchBot uses gen erative AI to rank potential matches and considers household and family information alongside individual demographic data. Case managers then review potential matches through an improved user interface. By strengthening data matching across SUN Bucks, the Supplemental

Participants of the Data SAIL kick-off meeting in Austin, TX, in February 2026.

Nutrition Assistance Program, Temporary Cash Assistance, Medicaid, Foster Care, and Free and Reduced Price Meals, Maryland aims to save caseworker time and reduce errors while supporting timely and accurate benefit delivery. Nebraska: Bridging Eligibility and Workforce Systems Nebraska is using Data SAIL to lay the groundwork for connecting NFOCUS, its eligibility system, with NEworks, its workforce case manage ment platform. Currently, the systems largely operate separately, requiring work force staff to duplicate participation information or exchange additional details through email. Nebraska is designing a pathway that would allow eligibility staff to access work force participation information more directly, reducing duplicative work

and improving coordination across programs. Rather than immediately beginning a technical build, Nebraska is focusing on discovery and planning. The state is mapping workflows, defining nec essary data elements, documenting requirements, and engaging staff and customers through focus groups and listening sessions. This work will also help prepare Nebraska for the planned integration of Temporary Assistance for Needy Families Employment First (TANF EF) into NEworks. Kentucky, Maryland, and Nebraska are approaching integration from dif ferent starting points, but their projects demonstrate a shared goal: designing services around people rather than individual programs and systems. By documenting lessons and sharing what states learn along the way, Data SAIL will help inform broader efforts to build more integrated, responsive, and customer-centered health, human services, and workforce systems across the country. Building a More Connected Future

By documenting lessons and sharing what states learn along the way, Data SAIL will help inform broader efforts to build more integrated, responsive, and customer centered health, human services, and workforce systems across the country.

Nicole Acosta is the Process Innovation Coordinator at APHSA.

11

Fall 2026 Policy & Practice

ai in focus

By David Turner

Laying the Groundwork for AI Involvement: Why Data Governance Is the Ultimate Prerequisite

H uman Services agencies provide a vital safety net for millions of people. They manage food assistance programs. They oversee statewide health care access. They provide child welfare initiatives and emergency support. These organizations, however, face immense pressure. Though often lumped together and summarized as simply needing to “do more with less,” the challenges facing these agencies today are many. The most notable of these include workforce shortages, changing eligibility requirements, and legacy technology. In isolation, these hurdles are serious enough; in combination, difficulties might seem exponential. While artificial intelligence (AI) has emerged as a potential solution—one that might allow safety net systems to operate more efficiently, reduce caseworker burnout, or dramatically improve constituent experiences— its value to organizations is limited without proper data governance. How Is AI driving Operational Efficiency? As it has in other industries, AI may be used in assisting state human Multilingual AI agents might allow state residents to engage through a more preferred channel (phone, chat, or web) and receive faster responses from applicants. Automated pro cesses might allow for quicker access to public health information, status checks, or application completion. services agencies’ attempts to: n Simplify User Experience:

Nutrition Assistance Program and Medicaid work verification efforts. n New Jersey: Through the New Jersey Innovation Authority, the State of New Jersey launched the NJ AI Assistant (Assistant) in 2024. 2 Since then, the state says agencies have used the tool to analyze taxpayer feedback and have accordingly adjusted certain automated offerings based on that analysis—resulting in an increase of self-resolved issues. New Jersey says it updated the Assistant’s infrastructure and AI model as recently as March 2026. Why Is Data Governance the Ultimate Prerequisite for Possible Government AI Tools? AI models are only as good as the data feeding them and the rules

n Boost Staff Efficiency: Caseworkers, no longer burdened with repetitive administrative work (file and document summary, meeting preparation), may find time to tackle secondary, nuanced, or more attention-demanding tasks. n Advance Analysis and Mitigation: AI can cross-check submitted appli cation data against state records in real time, possibly helping detect anomalies and mitigating risks.

States Moving From Theoretical to Operational n Maryland: The Maryland

Governor’s Office reports 1 plans to experiment with AI as a document processing tool. Additionally, the state says multiple agencies plan to use AI for various Supplemental

Illustrations by Chris Campbell

12

Policy & Practice Fall 2026

Robust data governance policies help ensure that the data that AI uses is more informed, consistently applied or reviewed, securely managed, and considerate of policy requirements.

putting boundaries around tasks. Should weak data governance result in poorly sourced AI workflows— especially in human services—there could be an impact on the provision of an agency’s services and, thereby, an impact on potential recipients, case workers, and taxpayers. Robust data governance policies help ensure that the data that AI uses is more informed, consistently applied or reviewed, securely managed, and considerate of policy require ments. State agencies might achieve this by establishing formal gover nance structures or innovation labs designed to continuously test systems for performance, error patterns, and operational compliance. In light of the sensitive data they handle, state agencies are regularly tasked with meeting stricter security benchmarks. This makes data conscious partnerships a must. Look for partners with certifications like “FedRAMP Ready” status or NIST or How Can Agencies Account for Data Security?

FISMA benchmarks, which signify that a data architecture has the necessary guardrails to protect consumer infor mation against escalating cyber threats and other similar considerations. For further examples, see Equifax’s security annual report, which high lights the importance of pairing advanced data analytics with a secu rity-focused approach. 3 The Path Forward: Moving From Adoption to Accountability As state human services agencies continue to integrate AI into their workflows, the long-term measure of success will be whether these tools seem to create a more responsive and human-centric process for seeking

government services. By establishing rigorous data governance, partnering with secure infrastructure providers, and keeping humans (both beneficia ries and caseworkers) at the center of their decision making, states can more responsibly work to deliver the true promise of AI. David Turner is the Senior Vice President and General Manager of Government Solutions at Equifax. Reference Notes 1. https://governor.maryland.gov/news/ press-releases/maryland-secures-ai-grants strengthen-delivery-public-services 2. https://innovation.nj.gov/projects/ ai-assistant/ 3. https://www.equifax.com/about-equifax/ security/annual-report

    ◦  ◦    ◦  ‚„‡ ˆ ‡  THE COMPASS

◦ € ‚  ◦ ƒ „  ‚ ◦  ‚ † 

13

Fall 2026 Policy & Practice

ai in focus

By Heidi Reed and Lauren Aaronson

A New Path Forward for Eligibility Modernization: How AI-Assisted Development, Applied with Deep Expertise, Can Make Modular Modernization a Reality

E very day, state eligibility systems decide who gets health coverage, food assistance, cash assistance, child care benefits, and more. They process millions of cases across dozens of programs, applying layers of federal and state rules at a scale that few other government systems match. Keeping them current has always been difficult, and the pressure to modernize has never been greater. States have long faced a difficult choice: keep patching aging systems that can barely keep up with policy change or pursue full replacements that carry enormous risk and cost. A third option is gaining traction. AI-assisted development tools, used by teams with the right expertise, are making modular modernization both practical and affordable. Complexity does not go away, but skilled teams now have better instruments to work through it. A typical state manages hundreds of eligibility rules across Medicaid, the Supplemental Food Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), and other programs, each governed by its own federal framework with distinct income definitions and verifi cation requirements. Those rules shift by program, state, county, and some times, household type. The complexity is an accurate reflection of the under lying policy, not a design flaw that modernization will solve. The Real Complexity States Are Managing

Teams also face a constant stream of policy changes (federal rule updates, state mandates, court orders) each requiring system modifications while the current platform continues serving people every day. Carving out space for strategic modernization inside that workload is one of the central chal lenges. The aim is to build systems that manage the complexity better, working through it incrementally rather than pretending it can be engi neered away. Why Modular Modernization, and Why Now The ability to divide a large system into smaller, independently deploy able parts has long been appealing in theory. Executing it is a different matter. You cannot simply separate one part of a legacy system and call it a module. Data move between

components, and business rules are often shared across the system in ways that are not well documented or fully understood. Getting the boundaries right requires people who understand both the technical architecture and the policy it encodes, and that combi nation is genuinely hard to find. Data migration makes this even harder. Moving a capability into a new module often means restructuring data, preserving years of historical records for audits and appeals, and keeping both old and new systems in sync during the transition. Errors in this process have real consequences for real people, so the work demands exceptional care. The approach follows a clear pattern: build new capabilities along side the existing system, route work to new modules gradually, and retire legacy components only after replace ments are proven in production. The

14

Policy & Practice Fall 2026

The economics favor a portfolio of incremental investments over high-stakes, all-or-nothing replacements. Shared components and standards become more viable, reducing duplicated effort across programs and states.

risk profile is far more manageable, and agencies see value at each stage rather than waiting years for a single delivery. What AI-Assisted Development Actually Adds Modular modernization has stalled in the past, less because of flawed strategy than because of the economics of execution. AI-assisted development is starting to shift that equation. Generative coding tools, in the hands of skilled developers, can draft inte gration code, generate test cases, and produce documentation in a fraction of the time these tasks used to require. The same tools help experienced teams thoroughly assess legacy systems: surfacing hidden logic, discovering undocumented dependencies, and defining how data truly flow through the system. Identifying these complexi ties early helps shape a modernization road map that delivers incremental value while reducing pressure on state staff. Used well, these tools can raise the bar on what a knowledgeable team can accomplish. The expertise still has to be there, but AI makes it go further. The Modular Rules Engine: Where AI and Formal Verification Converge Of all the components in an eli gibility system, the rules engine is the one most directly responsible for correctness, consistency, and compli ance. A typical mid-size state manages hundreds of interdependent eligibility rules across programs, with federal and state versions that must stay syn chronized. Changes to one program’s criteria frequently cascade across others. Manual processes struggle to trace those dependencies safely, and bugs discovered weeks into production can cause systematic overpayments or wrongful denials affecting thousands of applicants before they are detected. A modular rules engine, one that encodes income calculations, house hold composition logic, categorical eligibility pathways, and other policy rules in discrete, testable, independently updatable modules, is a foundational asset for any modern benefits platform. It does not

determination is backed by a formal proof chain that provides defensible evidence for federal compliance reviews and applicant appeals. For states, this means the rules engine can finally keep pace with the velocity of federal policy change, without the risk of silent errors accumulating in production.

eliminate the interconnectedness of eligibility logic, but it makes that interconnectedness explicit and man ageable. When federal policy changes, the affected rule modules can be iden tified, updated, tested, and deployed without requiring changes to unre lated parts of the system. Accenture and AWS have access to capabilities that take this further than traditional rules engine approaches. AWS Automated Reasoning, available through Amazon Bedrock Guardrails, uses formal mathematical verification to prove that eligibility decisions are correct, not just that rules are executed as coded. When a new federal rule is introduced, the system can instantly verify whether it conflicts with existing state rules across all programs before deployment, rather than discovering contradictions weeks into production. Rule changes that once required two to four weeks of manual specification, coding, and regression testing can be validated and deployed in one to two days, with formal proof of correctness serving as an audit artifact. This matters because eligibility determinations require deterministic, auditable logic, not probabilistic infer ence. Agentic AI is a powerful tool for adjacent tasks: document processing, data ingestion, case quality review, case triage, and derived insights. But the core rules engine demands certainty. The combination of large language models (LLMs) for inter preting regulatory language and Automated Reasoning for formal verification delivers both: AI’s ability to rapidly translate federal guidance into structured rule specifications, backed by mathematical proof that those specifications are logically consistent and cross-program compliant. Every

A Moment Worth Taking Seriously When modular architecture,

AI-assisted development, and formal verification work together, the relation ship between policy and technology starts to shift. Policy teams can see their intent reflected in systems within a realistic time frame. The economics favor a portfolio of incremental invest ments over high-stakes, all-or-nothing replacements. Shared components and standards become more viable, reducing duplicated effort across programs and states. States pursuing this path need partners who bring both technical depth and policy knowledge. AI tools can accelerate the work considerably, but decisions about where and how to apply them depend on domain knowl edge that takes years to develop. None of this is simple, and the expertise required is substantial. For states willing to invest in under standing their architecture before changing it, the tools are now avail able to make modular modernization an achievable idea.

Heidi Reed is the Integrated Eligibility Lead at Accenture.

Lauren Aaronson is the HHS Technology and Innovation Lead at Amazon Web Services (AWS).

15

Fall 2026 Policy & Practice

STRENGTHENING SNAP PERFORMANCE TODAY

Building the Foundation for Modernization

By David Crowson and Pat Aguilar

tate human services agencies administering the Supplemental Nutrition Assistance Program (SNAP) are operating in an increasingly complex and high-stakes environment. Expectations for accuracy, timeliness, and program integrity continue to rise, while workforce constraints, increasing caseloads, and new federal requirements place sus tained pressure on program performance. These pressures are not theoretical. Recent federal policy changes have heightened the financial implications of elevated payment error rates, increasing cost-share exposure for states and reinforcing the need to get eligibility determinations right the first time. At the same time, agencies must continue to ensure that SNAP benefits are delivered accurately and efficiently. SNAP leaders don’t have to choose between performance improvements and modernization. Instead, they can pursue both in tandem through a coordinated strategy that aligns oper ations, technology, and policy. Agencies that connect today’s priorities with a comprehensive vision for modernization can strengthen program integrity, reduce workforce strain, and build more resilient systems for the future.

17

Fall 2026 Policy & Practice

The Challenge of DeliveringToday While Building for the Future Many SNAP agencies are cur rently navigating a dual imperative: maintain or improve compliance and operational stability today while planning for a more modern, efficient, and resilient future. The reality is that states need to prioritize near-term actions to address performance challenges, including reducing payment error rates, stream lining eligibility determinations to improve timeliness, and managing workload pressures. These targeted interventions play an essential role in sustaining operations and mitigating financial risk. At the same time, achieving longer-term progress may require a modernization plan that connects near-term performance priorities with a defined future-state vision. This plan doesn’t need to be at odds with today’s needs. By synchronizing immediate improvements with a defined mod ernization plan, states can improve performance and operational stability now while building the foundation for incremental enhancements or broader system transformation later. A structured approach—stabilize, redesign, and scale—offers a practical way to sequence efforts and avoid the

tradeoffs that often undermine both near-term performance improvements, and process and system modernization. Stabilize First by Strengthening Accuracy and Capacity Start by stabilizing operations and improving accuracy at key decision points. Traditionally, many agencies have relied on downstream quality assurance processes to identify and correct errors after eligibility determi nations have been made. A more effective strategy is a pro active approach to error prevention: catching errors before benefits are issued, flagging missing or conflicting information early, and guiding case workers through complex decisions. proactive approach to error prevention: n Catching errors before benefits are issued n Flagging missing or conflicting information early n Guiding caseworkers through complex decisions By embedding upstream quality assurance processes and technology improvements directly into eligibility workflows, data inconsistencies within cases can be identified and resolved before benefits are issued. Technology enabled, human-in-the-loop tools can reduce staff burden and improve accuracy at the point of action. Additionally, by focusing on continuous case confidence rather than just case completion, agencies can better assess readiness, route work appropriately, and target supervision where the risk of error is greatest. Stabilization also requires addressing staff capacity constraints. Skilled workforce shortages and fluctuating caseloads strain staff and contribute to processing delays and errors. Blended workforce models that combine state staff, contractors, and automation offer the flexibility to manage surges. SHIFT FROM CASE COMPLETION TO CASE CONFIDENCE A more effective strategy is a

Agencies can deploy surge support to stabilize operations during peak demand or policy-driven workload spikes. When effectively integrated, this approach combines staff augmen tation with technology and process support to maintain or improve performance standards without over burdening existing teams. By focusing on proactive accuracy, upstream quality assurance, and flexible staff capacity, agencies can reduce rework, improve consistency, and create a more stable operational baseline. This stability is essential to enable the next phase of transforma tion and ensure modernization efforts succeed without introducing new operational risks. Redesign Next by Reworking Processes to Enable Modernization With operations stabilized and vari ability reduced, agencies can turn to redesign, reworking processes and workflows to support long-term modernization. It’s key to remember that moderniza tion alone cannot solve deeply rooted operational challenges. Without first addressing how work gets done, sig nificant technology upgrades risk automating inefficient processes or reinforcing existing silos. Business process reengineering—informed by real-time data and the knowledge of where and when bottlenecks and errors occur—provides a structured approach to redesigning workflows, staffing models, and decision points to improve efficiency and consistency. This does not always require replacing or heavily modifying core eligibility systems, which can be slow, costly, and difficult to change. Instead, agencies can introduce focused deci sion-support capabilities that work alongside existing eligibility systems to help guide staff decision making and streamline workflows. Key operational components often present the best opportunities for improvement: n Complexity-based routing and task assignment can ensure eligibility tasks are directed to the appropriate staff with supporting technology, reducing bottlenecks.

David Crowson is the Senior Vice President of Business Solutions for Maximus.

Pat Aguilar is the Managing Director of Program Modernization for Maximus.

18

Policy & Practice Fall 2026

Scaling modernization doesn’t require a one-size-fits-all approach. Some agencies may pursue incre mental enhancements, building on existing systems and processes over time, while others may undertake broader system transformation initia tives. The key is to synchronize the pace and scope of change with opera tional readiness and institutionalize performance management practices that sustain improvements over time. Doing so ensures that tech nology investments are effectively adopted, integrated, and sustained. It maximizes impact while enabling improved performance and reducing the risk of disruption. KeyTakeaways for SNAP Leaders For state human services leaders, several key principles emerge from this approach: n Connect near-term performance priorities with a defined future-state vision to build the foundation for modernization success. n Use human-centered, technology enabled tools to guide eligibility decisions and strengthen accuracy at the point of action. function, using QA findings to inform staff coaching, workflow updates, and risk-based routing rules, enabling agencies to continuously learn from errors and prevent them over time. LEVERAGE SMART TOOLS FOR CONTINUOUS IMPROVEMENT Decision-support tools, including those that use artificial intelligence, can flag risk factors, summarize key case facts, and provide relevant guidance to caseworkers. When combined with human oversight, these tools can strengthen accuracy while maintaining accountability and program integrity, and they can be scaled to drive more consistent performance. These tools can be added to your existing eligibility ecosystem without requiring system replacements. Quality assurance (QA) can also become a closed-loop learning

n Redesign processes and workforce models before scaling technology, ensuring that modernization investments are built on a strong operational foundation that rein forces efficient workflows. n Measure operational burden as part of performance management by examining verification burden, rework loops, policy lookups, handoffs, and exception decisions that add unnecessary complexity. n Keep staff experience and customer experience central to all moderniza tion efforts, recognizing that both are critical to successful outcomes. improvement and modernization as separate efforts to managing them as interconnected components of a single, coordinated strategy. Why It Matters The stakes for SNAP agencies are high. Financial exposure from payment errors, increased workforce strain, and risks to public confidence and trust underscore the importance of effective program management. At the same time, the opportunity is significant. Agencies that take a coordinated, staged approach to improvement can achieve meaningful gains in program integrity, workflow efficiency, and operational resilience. Immediate performance gains, such as reduced error rates and improved time liness, create a strong foundation for future modernization. By stabilizing operations, redesigning processes, and scaling technology in agreement with a long-term strategy, agencies can build systems that are more reliable, efficient, and responsive. These systems are better equipped to meet the needs of individuals and families while adapting to evolving policy and operational demands. Ultimately, transforming SNAP delivery is about connecting immediate performance improvement with long term innovation so that today’s actions support tomorrow’s outcomes. Through a structured, holistic approach, agencies can not only navigate current challenges, but also build a more resilient and modern future for human services. Together, these principles reflect a critical shift from treating performance

n Verification burden management can help agencies identify duplicative or difficult-to-obtain documentation, time spent resolving issues, and where available data may reduce unnecessary follow-up. n Error pathway mapping can help agencies understand when and how errors develop across the life of a case, enabling earlier intervention. n Case flow, decision points, and handoffs can be decomposed and reengineered to minimize unneces sary steps, clarify responsibilities, and reduce delays. n Policy-to-process traceability can create a clear line of sight from policy requirements to system fields, workflows, quality checks, and training materials, reducing the risk of operational gaps when policies change. Effective reengineering efforts are grounded in a clear understanding of both staff and customer experience. This means examining how work gets done while also improving how indi viduals and families interact with the program. Ensuring alignment between the two perspectives creates more effi cient, accessible processes. Process redesign should precede large-scale technology investments. Streamlined, well-defined business processes provide a stronger foun dation for system enhancements, ensuring that technology support, rather than complicate, operations. Then Scale While Aligning Strategy, Operations, and Technology The final step is to scale modern ization efforts by expanding and standardizing successful practices across operations to align strategy, operations, and technology. Once processes are stabilized and redesigned, agencies can further leverage data analysis and advanced tools to drive continuous process improvement. Data play a critical role in supporting decision making, per formance management, and program oversight. It can be used to track error patterns, monitor timeliness across applications and recertifications, and identify verification bottlenecks at scale.

19

Fall 2026 Policy & Practice