The H.R. 1 Windfall: Vendors, AI, and Safety Net Access

• In Event: Administrative Burdens and Effectiveness of the Social Safety Net

Abstract

CONTEXT: The passage of H.R. 1 in July 2025 imposed significant new administrative burdens on individuals seeking to access the two largest safety-net programs: Medicaid and the Supplemental Nutrition Assistance Program (SNAP). For the first time, H.R. 1 mandates work requirements for adults ages 19 to 64 without disabilities who receive Medicaid. In addition, this same enrollee group will now be required to renew their Medicaid coverage every 6 months, instead of each year. For SNAP, work requirements were also expanded significantly, while the ability of states to seek waivers from these requirements based on economic conditions was tightened. Although these changes are estimated to reduce Medicaid coverage by about 6 million and SNAP by 2.4 million, they also represent substantial revenue generating opportunities for companies that provide state governments with program eligibility, compliance, and income verification services. Many states have made clear that they will rely heavily on their vendors to respond to H.R. 1., while vendors in turn have stated that they will ramp up use of A.I. to support states as they make the urgent changes needed for compliance. 

METHODS: This paper includes two elements: 1) a vendor landscape analysis to identify the main vendors that states are likely to use in Medicaid and SNAP for H.R. 1 response and 2) an analysis of earnings call and press release data from the major vendors to assess how they describe changes to their revenue and use of A.I. The former part will review the major vendors that currently or plan to provide services for Medicaid and SNAP in the following areas: income and employment verification, eligibility and enrollment systems, call centers and member services, Medicaid management information systems, and SNAP employment and training systems. The first three functions are often shared between programs, while the last two are program specific. It will map the services that each vendor provides, identify which states the vendor provides these services, collect data on the revenue streams that each vendor receives across services, and measure A.I. use or planned use. In addition, given that individual states have imposed requirements similar to those now mandated by H.R. 1, the landscape analysis will provide background research on vendor involvement in a few example state-level changes. In the second component, we will collect earnings call data from CapitalIQ and scrape press release data from company websites. Then, manual review of transcripts and press releases will initially be performed to create coding categories. NPL will then be used to analyze the full dataset using these categories. 

RESULTS: We will identify the main industry players and how each provides services across the range of state needs. In addition, we will be able to characterize how vendors are selling A.I. use in response to H.R. 1 and which states are buying these services. Ultimately, we can estimate anticipated earnings for these companies as they work with states to respond to H.R. 1., provide administrative cost estimates, and identify areas where state purchases could be more efficient.

Author

• Jeremy Barofsky, Georgetown University
  Presenting Author