The Human Advocacy Layer

The organizations covered in this article are not distribution channels in the commercial sense. They are not positioned to drive technology adoption through enterprise sales cycles, franchise licensing agreements, or pharmacy chain procurement committees. What they are doing is something more important and structurally different: they are already providing, manually and at insufficient scale, exactly the navigation and advocacy services that the technology sector is attempting to automate. SHIP counselors compare Medicare plans one-on-one. ADRC specialists screen for benefit eligibility across seven or more programs simultaneously. AAA case managers walk seniors through SNAP recertification and MSP enrollment. Benefits enrollment organizations file applications for programs eligible seniors have never heard of.

The technology sector’s relationship with these organizations is therefore not a go-to-market partnership in the standard commercial sense. It is a service delivery collaboration — AI handles information synthesis and administrative preparation at scale; human advocates handle judgment, exceptions, execution, relationship, and the bureaucratic interventions that software categorically cannot complete. A SHIP counselor who arrives at a counseling session with an AI-generated cross-program eligibility profile, a formulary change summary, and a prior denial explanation already drafted is more effective than one who builds that picture from scratch during a 45-minute appointment. The AI does not replace the counselor. It changes what the counselor can accomplish in the time she has.

This distinction changes what gets built. A product designed for commercial distribution channels needs a beneficiary-facing mobile interface and an administrator dashboard. A product designed for integration with the human advocacy layer needs a professional case management interface that allows counselors, SHIP volunteers, and AAA case managers to access a client’s benefit landscape, act on flagged items, record completed steps, and document what remains. These are different products. Organizations building in this space eventually need both.

Aging and Disability Resource Centers

#

ADRCs are a collaborative program of the Administration for Community Living, CMS, and the Veterans Health Administration, operating as single points of entry into the long-term services and supports system for adults of all income levels and all disability types. The design principle — No Wrong Door — reflects the recognition that the population seeking LTSS cannot be expected to navigate the correct bureaucratic entry point before receiving help. Any door should be the right door.

ADRC staff are benefit specialists whose job is precisely the cross-system coordination problem that MCR-06.12 identifies as the core burden: Medicare, Medicaid, Social Security, SNAP, LIHEAP, state pharmaceutical assistance programs, LTSS waiver applications, and community social services. The ADRC is the institutional infrastructure designed to hold all of that simultaneously and guide an individual through it. As of FY2024, ACL’s discretionary funding to ADRCs was $8.6 million nationally — a figure that reflects the structural underfunding of an institution whose mandate has grown substantially faster than its budget. The 1,322 access points nationwide, which include local AAAs, Centers for Independent Living, and tribal organizations alongside dedicated ADRC offices, represent the geographic reach of the No Wrong Door system. The resource base behind that reach is thin.

The capacity constraint is operational, not motivated. ADRC counselors carry caseloads that make proactive outreach impossible. The modal ADRC contact is reactive — a senior in crisis, a caregiver at the breaking point, a family arriving at an unfamiliar system after a hospitalization. Early identification of benefit gaps for seniors who are not yet in crisis — the preventive use case that would generate the most value — requires outreach capacity that most ADRCs do not have. An AI tool that proactively surfaces eligibility gaps and prepares preliminary documentation, then routes the flagged case to an ADRC counselor for confirmation and action, converts the ADRC from a reactive crisis responder into a proactive benefit access partner. The counselor’s judgment and execution capacity remains the bottleneck. The tool removes the information synthesis steps that currently consume the time before the judgment begins.

The No Wrong Door philosophy also defines the scope of what an AI navigation tool deployed through ADRCs must cover. ADRCs serve individuals at all income levels, not just those below Medicaid eligibility thresholds. A tool that screens only for income-based programs misses the middle-income senior who qualifies for State Pharmaceutical Assistance Programs but not Medicaid, the homeowner who qualifies for a property tax freeze program for seniors but not SNAP, and the veteran who qualifies for VA benefits but has never applied because no one told her she was eligible. Universal eligibility screening — not means-tested program lookup — is the standard the ADRC system operates at and the standard an AI navigation tool must meet to be useful in that context.

SHIP Programs and the Benefits Enrollment Ecosystem

#

SHIP is a national program operating through 54 grantees — one in each state, the District of Columbia, Puerto Rico, Guam, and the U.S. Virgin Islands — with more than 12,500 team members including staff, in-kind professionals, and volunteers across 2,200 local sites. SHIP is funded through ACL’s Office of Healthcare Information and Counseling and operates through state units on aging and state departments of insurance in partnership with local AAAs. It is the most trusted source of free Medicare counseling for beneficiaries, and its counselors are trained and certified to advise on Original Medicare, Medicare Advantage, Part D, Medigap, MSP enrollment, and LIS. SHIP also assists with fraud identification and beneficiary rights — the scope is broad because the Medicare system’s complexity requires it to be.

The bandwidth problem is structural. SHIP operates on the assumption that one-on-one counseling from a trained expert is what Medicare beneficiaries need, which is correct. The number of trained experts relative to the number of beneficiaries who need them is the constraint that no SHIP budget has ever adequately addressed. A SHIP counselor handling three to four complex cases per day with manual research tools is operating at or near the practical capacity of what a human can do in a six-hour counseling day — plan comparison, benefit screening, document preparation, application assistance, and follow-up. That same counselor, with AI pre-work completed before each session — an eligibility profile, a plan comparison, a prior denial explanation, a documentation checklist — could handle more cases, handle them more thoroughly, or spend the session time on the edge cases that actually require her expertise rather than on information gathering that AI can do faster and more comprehensively.

The benefits enrollment organizations that complement SHIP provide application assistance that SHIP’s counseling model is not always structured to complete. The Medicare Rights Center operates a national helpline that answered thousands of MSP and Medicare benefit questions in 2024 and documents the application barriers beneficiaries face as a policy advocacy function. The National Council on Aging’s Center for Benefits Access operates BenefitsCheckUp, the existing digital benchmark for cross-program benefits screening. BenefitsCheckUp is a point-in-time lookup tool that identifies eligibility based on user-entered income and household information and generates a list of programs the user may qualify for. It does not provide longitudinal tracking, proactive alerts when eligibility changes or deadlines approach, conversational guidance for users who do not know what questions to ask, or integration with the case management workflow of the counselors who follow up after the screening. It is the best existing tool in this space and the starting point against which any new entrant must differentiate.

The integration model for AI tools working alongside SHIP and benefits enrollment organizations is straightforward in concept and difficult in execution. AI identifies eligibility signals from the information a senior provides and surfaces them to the counselor or specialist for confirmation and action. The counselor’s expertise is applied to the cases that require it — the appeals with legal nuance, the exceptions processes requiring human-to-human intervention, the beneficiaries whose documentation situation requires judgment about how to proceed — rather than to the initial screening and information gathering that could be handled upstream. The execution difficulty is workflow integration: SHIP and BEO counselors work in state and local technology systems that are not designed to receive AI-generated eligibility profiles. The professional interface has to connect to those systems or operate effectively alongside them, and each state’s SHIP infrastructure is different.

Area Agencies on Aging

#

The OAA network, which the Supporting Older Americans Act of 2020 reauthorized through FY2024, operates through 56 State Units on Aging and 618 Area Agencies on Aging, serving 10.1 million older persons through Title III programs as of the most recent comprehensive reporting year. In FY2025, OAA programs received $2.372 billion in funding — $392 million below the authorized level, a gap that the National Association of Counties and aging advocacy organizations have consistently flagged as the primary constraint on expanding service capacity to meet demographic demand. AAAs provide information and referral, home-delivered and congregate meals, transportation, caregiver support, benefits counseling, and evidence-based health promotion under the OAA Title III structure.

The AAA network’s geographic reach — covering every county in the country — is an asset that no technology platform can replicate, particularly in rural communities where the AAA may be the only organized aging services infrastructure present. The AAA has the community relationship and the beneficiary trust. Technology can extend what the AAA can do with those relationships; it cannot substitute for them in communities where the institutional presence of the AAA is what makes any service delivery possible.

OAA Title III funding creates natural entry points for digital tools. Title III-B supportive services funding covers information and referral — the service category most directly aligned with AI-powered navigation tools. Title III-D evidence-based disease prevention and health promotion funding has an established review process, reopened in January 2025 under NCOA leadership, for identifying programs with demonstrated evidence of improving health and well-being among older adults. A navigation or benefit-access tool with published deployment data and documented outcome improvements can apply for inclusion in the ACL evidence-based program list — inclusion that makes the tool eligible for Title III-D funding through state aging plans and AAA subgrants.

The SHIP funding pass-through creates a specific mechanism. SHIP grants flow from ACL through state units on aging, which frequently pass funds through to AAAs and other local partners to provide SHIP services at the community level. Technology tools that support Medicare counseling activities — plan comparison, benefit eligibility screening, documentation preparation — can be incorporated into SHIP subgrant activities if state SHIP grantees choose to include them. The tool needs to serve the SHIP function, not just exist alongside it. A SHIP subgrantee that incorporates AI-assisted benefit screening into its counseling workflow has a legitimate basis for including the tool cost in its SHIP budget.

The ACL OAA final rule that took effect March 15, 2024, with a compliance date of October 1, 2025, modernized OAA program regulations for the first time since 1988. Among the changes: requirements around equity in service delivery, accountability for funds expended, and explicit priority for serving those with the greatest economic need and greatest social need — particularly low-income minority older individuals, those with limited English proficiency, those in rural areas, and those with disabilities. These priority populations are exactly the populations for whom the cognitive burden described in MCR-06.12 is most severe and for whom AI navigation tools, if designed with appropriate accessibility and language support, provide the most differentiated value.

Meals on Wheels and the Homebound Population

#

Meals on Wheels America serves approximately 2.4 million seniors annually through a network of more than 5,000 local programs. The population it serves — homebound seniors, many of whom are socially isolated, many of whom have limited mobility or cognitive impairment — is the hardest-to-reach population for any technology platform that depends on active user adoption. A homebound 89-year-old with macular degeneration and mild cognitive impairment is not going to download an app from an app store.

The meal delivery volunteer is one of the few people who has consistent, regular, in-person contact with this population. In many communities, the Meals on Wheels volunteer is the only person who sees a particular senior in person during a given week. That contact creates the trust and the access that make technology introduction possible — not through a digital channel, but through the human relationship that already exists. A volunteer who delivers meals and introduces a tablet-based companion tool, shows the senior how to use it, and mentions that it can help her understand her Medicare plan is providing a service that no digital onboarding flow can replicate.

The food insecurity and healthcare connection is not incidental. Food insecurity is a documented driver of medication non-adherence — a beneficiary choosing between groceries and prescriptions is making a choice that produces care plan failures, hospitalizations, and Medicare spending that every CMMI model is designed to prevent. The Meals on Wheels visit is an informal clinical touchpoint whether it functions as one explicitly or not. Home-delivered meals programs that incorporate basic wellbeing checks — is she taking her medications, does she seem confused, are there new safety concerns — are already operating at the boundary between social service and clinical outreach. A tool that the volunteer can use to flag concerning observations for follow-up by the AAA case manager or the home health agency extends that informal clinical function without requiring volunteers to perform clinical assessments they are not trained to conduct.

What This Means for Product Design and Business Model

#

The product design implication from the advocacy layer is specific and non-negotiable: an AI navigation platform that intends to serve the highest-need Medicare population requires a professional interface. The beneficiary-facing interface handles the senior who engages directly — capable older adults who can use a tablet or smartphone, family caregivers supporting someone who cannot. The professional interface handles the ADRC counselor, the SHIP volunteer, the AAA case manager, and the Meals on Wheels coordinator who needs to know what the senior she visited yesterday may be eligible for, what is outstanding from the last appointment, and what action is needed before a deadline next month.

Without the professional interface, these organizations cannot incorporate the tool into their workflows, and the seniors who need it most — the homebound, the cognitively impaired, the isolated, the linguistically isolated — remain unreached. The commercial distribution channels in MCR-06.13 serve seniors who can engage with technology directly. The human advocacy layer is the only channel that reaches the seniors who cannot.

The business model implication follows the product design. The advocacy organizations in this article are not primarily revenue-generating partners in the near term. The SHIP program operates on a federal grant. The ADRC system is funded through ACL discretionary appropriations totaling $8.6 million nationally — a number that leaves no procurement budget for enterprise technology licensing. AAAs operate on OAA formula grants with limited discretionary spend. Meals on Wheels programs are predominantly volunteer-driven nonprofits.

The value of these relationships is not near-term revenue. It is the credibility infrastructure that validates efficacy with the highest-need population and produces the outcome data — benefit enrollment rates, administrative burden reductions, case resolution times, beneficiary-reported wellbeing measures — that government and commercial payers eventually require as the basis for any reimbursement or coverage decision. The NYSOFA-ElliQ model described in MCR-06.10 demonstrates what this looks like in practice: a state government funds the deployment, the local aging agencies handle identification and installation, the data flows back to the developer and the funder, and the published outcomes become the evidence base for the next funding cycle and the next expansion.

The AAA and ADRC network is the evidence-generation channel. The commercial channels in MCR-06.13 are the revenue channel. The sequencing of the business model requires both, and it requires the advocacy-layer relationships to come first — because without the outcome data that serving the highest-need population generates, the commercial case to payers and operators is built on market projections rather than demonstrated impact.

Related Reading

#

MCR-00_03 The Medigap Market MCR-10_01 The LIS Landscape: Extra Help, Medicare Savings Programs, and the Low-Income Non-Dual Population MCR-11_04 Arizona and Nevada: Sun Belt Medicare in the WISeR Era