{"resourceId":"nys-doh-converged-ee-rfp-ai-contract-terms-20261029","versions":[{"version":"external-96bc37eae27c53af2df40e7b476f9be19fe52993e59f9ce859d30c10ba690018","resource":{"id":"nys-doh-converged-ee-rfp-ai-contract-terms-20261029","title":"New York DOH's seven-year Medicaid eligibility system RFP asks vendors to use AI, but every use needs 90 days' notice and written approval, with non-essential AI off by default","organization":"New York State Department of Health (Office of Health Insurance Programs)","sector":"State Medicaid and health insurance eligibility and enrollment","geography":"New York, United States","publishedAt":"2026-10-29","publicationDate":"2026-10-29","eventDate":"2026-11-13","sourceName":"NYS DOH RFP C043776: Converged Eligibility and Enrollment (E&E) System and Appeals Management System (AMS)","sourceLabel":"State request for proposals (PDF)","sourceUrl":"https://healthweb-back.health.ny.gov/funding/rfp/c043776/rfp.pdf","evidenceClass":"standards-guidance","outcomeClass":"emerging","topics":["governance-procurement","data-security","developers-agents","infrastructure"],"finding":"The RFP, issued Oct. 29, 2026, seeks one contractor for a seven-year term. The contractor would maintain and operate New York's Medicaid Eligibility and Client Management (MECM) system and Appeals Management System. It would also fold NY State of Health functions into MECM, creating one Converged E&E System for Medicaid, Child Health Plus, the Basic Health Program and Qualified Health Plans. Section 5.2.9 encourages the contractor to use AI to improve quality and cut time and cost. The department's vision cites AI-enabled, model-driven solution engineering to lower costs. Section 6.10 controls that AI. The contractor must give 90 days' written notice and get written approval before using any AI. AI not integral to the system is disabled by default. The department may disallow any AI for any reason, and rejected AI must be replaced at no cost. Department data and AI outputs may not be used to train models. The department owns prompts and outputs. AI transcription is off unless requested. The department may ask for an independent audit or a NIST-defined AI impact assessment. Letters of intent are due Nov. 13, 2026; proposals Feb. 19, 2027.","sledRelevance":"The RFP covers 6.5 million enrollees across the programs (Table 2, July 31, 2026), so it is one of the largest state eligibility procurements in the country. It shows how a state Medicaid agency can encourage vendor AI for engineering productivity while keeping tight approval and data controls. Other states can reuse the clause set in their own eligibility contracts.","evidence":"Primary procurement document. It sets requirements and states expected outcomes; it reports no measured results from AI use.","architectureImplications":"The department's vision cites AI-enabled, model-driven engineering using low-code and no-code platforms and CI/CD. Today's eligibility engine runs on the open-source Drools rules engine, and the Appeals Management System runs on Salesforce. Each AI notice must describe the technology, its purpose, its connections to the department, its deployment timing and whether it is open source. Disabled-by-default AI means configuration control over AI features embedded in vendor platforms.","governanceImplications":"Gives the department written approval of every AI use, a right to disallow any AI for any reason, and audit, impact-assessment and explainability requests. It references NYS ITS policy NYS-P24-001 and an OHIP AI Security and Privacy Baseline (the baseline itself was not reviewed).","securityPrivacyImplications":"The RFP bars training on confidential data without the information owner's written approval and bars using prompts or outputs to train or tune models. The department owns AI outputs, and the contractor may not share department data with third parties.","caveats":"This is a solicitation, not an award or a deployed system. Requirements may change through Q&A (responses due on or about Jan. 19, 2027) and Attachment E, the requirements matrix, which was not reviewed. Expected AI cost savings are department aspirations. The 6.5 million total is NYSOH enrollment as of July 31, 2026, not the full Medicaid caseload.","streamIds":["health-human-services"],"roles":{"sales":"Sales framing: New York wants AI-assisted delivery on its Medicaid eligibility platform, but only with prior notice and written approval. Bidders and partners can offer a documented AI usage plan that fits Section 6.10. Lead with how AI is governed, not with feature counts. Do not promise AI savings or quick approvals; the RFP can disallow any AI for any reason and requires replacement at no cost.","engineering":"Engineering framing: design AI use as switchable components under department control. Each one should have documentation covering purpose, data connections, deployment timing and open-source status. Keep embedded AI and transcription off by default, block training on prompts and outputs, and prepare NIST-style impact assessments and explainability material. Separate AI used for code and test generation from AI that touches eligibility logic.","delivery":"Delivery framing: the department's Office of Health Insurance Programs owns approval; the contractor owns compliant use. Delivery plans should build in the 90-day notice window, approval tracking and fallback non-AI paths. Proposed acceptance checks are an approved AI register, evidence that disabled-by-default settings hold in production, and audit-ready logs. Track each approval request and its decision date so schedule risk shows up early. These are proposals, not RFP results."},"roleDetails":{"sales":{"customerProblem":"The state wants modern, faster eligibility delivery and lower vendor costs, while keeping strict control over any AI inside a Medicaid system.","stakeholders":["NYS DOH Office of Health Insurance Programs leadership","Procurement and contract management","NYS ITS and security/privacy officers","Medicaid eligibility operations","CMS certification reviewers"],"discoveryQuestions":["Which AI uses would the department approve first: engineering, testing, or operations?","What evidence satisfies the OHIP AI Security and Privacy Baseline?","How will the department judge whether AI is 'integral to the system'?","What does a no-cost replacement mean in pricing?"],"valueHypothesis":"Bounded hypothesis: a governed AI usage plan for development and testing can shorten delivery cycles without putting eligibility decisions under AI control.","potentialEngagement":"Proposal support or a partner role preparing the AI notice, approval and impact-assessment package under Section 6.10.","unsupportedClaims":["That AI will reduce costs by any specific amount","That any AI use is pre-approved","That the RFP authorizes AI eligibility decisions","That terms will not change during Q&A"]},"engineering":{"fit":"Fits AI-assisted engineering (code, test, documentation) and tightly scoped operational aids within state-controlled tenants.","architectureIntegration":"Switchable AI components in the MECM and Converged E&E stack, CI/CD and low-code tooling under department ownership, and audit logging for each approved use.","prerequisites":["Written department approval per use","90-day advance notice","Mapping to NYS-P24-001 and the OHIP AI baseline","Impact assessment and explainability material on request"],"constraints":["Non-integral AI disabled by default","No training on department data, prompts or outputs","AI transcription off unless requested","Department may disallow any AI at any time"],"security":"Department owns prompts, uploads and outputs; no third-party sharing; reports for third-party audits and an independent audit or NIST AI impact assessment may be requested.","proofOfValue":["Demonstrate AI features can be fully disabled without breaking functions","Show that prompts, uploads and outputs are never used to train or tune models","Provide a sample NIST-aligned impact assessment","Measure cycle-time change on a non-eligibility engineering task against a baseline"]},"delivery":{"work":"Run an AI governance track inside the contract: notice, approval, register, monitoring and fallbacks, alongside transition-in and convergence work.","dependencies":["Department approval capacity","Attachment E AI requirements","CMS Streamlined Modular Certification","NYS ITS policy updates"],"ownership":"NYS DOH OHIP owns approval and data; the selected contractor owns compliant AI use and evidence.","skillsAdoption":"Contractor and state staff need shared practices for AI notices, impact assessments and reviewing AI-generated code and documents.","governanceCheckpoints":["90-day notice before AI use","Written approval per tool","Periodic audit or impact assessment on request","Re-approval when AI features change"],"acceptanceCriteria":["Proposed: complete AI register matching approvals","Proposed: verified disabled-by-default configuration in each release","Proposed: zero use of department data for model training"],"risks":["Approval delays stall delivery schedules","Vendor platforms add AI features silently","Savings assumptions built into pricing fail","AI-generated code introduces defects into eligibility logic"]}},"retrievedAt":"2026-11-02T08:37:03-06:00","enrichedAt":"2026-11-02T08:42:25-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"The 90-day notice and approval duty also covers AI used in interactions with department staff, public users and data subjects. Users must be told about any functionality lost when transcription is disabled. The AI sections do not address effects on state eligibility staff.","procurementImplications":"If the department rejects an AI tool, the contractor must replace it with a non-AI solution or an approved alternative at no cost. If neither is acceptable, the department may terminate the contract at no cost. Bidders must price for that risk and for 90-day approval lead times. Expected savings from AI-assisted engineering are a stated department goal, not a contract guarantee.","operatingModelImplications":"The department's stated objective is to increase its ownership of licenses, code, configurations, documentation and data and to reduce reliance on vendors. AI is treated as a vendor productivity tool under state control, not as an autonomous eligibility decision-maker.","sourceVerification":{"openedUrl":"https://healthweb-back.health.ny.gov/funding/rfp/c043776/rfp.pdf","referenceExcerpt":"Any AI tools not integral to the system must be disabled by default. Implementation or use of AI must be approved in writing","promptVersion":"sled-research-v4.0 HHS nightly brief (backfill for 2026-11-01)","model":null,"basis":"agent-reported inspection"}}}]}