From the State Government edition of October 18, 2026
CMS (Oct 15) says it will soon publish results of its Indiana Medicaid AI prepayment-fraud pilot with Oracle and is running pilots with five more states, while conceding parts of its WISeR AI prior-authorization model are not working
Centers for Medicare & Medicaid Services (Deputy Administrator Kimberly Brandt); Indiana Family and Social Services Administration; Oracle · State Medicaid program integrity (federal-state AI fraud detection) · United States: Indiana pilot plus five unnamed states
- Publisher
- Nextgov/FCW
- Original publication
- 2026-10-15
- Source retrieved
- 2026-10-19
- Event date
- 2026-10-15
What happened
Speaking Oct 15 at the ADAPT conference, CMS Deputy Administrator Kimberly Brandt said CMS will soon publish results of the Indiana Medicaid fraud pilot it ran with Oracle and is actively working with five other states on Medicaid pilots. Per her slideshow, the pilots use AI to flag upcoding, phantom services and geographic anomalies in near real time before payment, and test a shared case platform for CMS, Medicaid Fraud Control Units, program integrity contractors, states and law enforcement; the stated goal, if proven, is a national standardized platform. No pilot results were reported. Brandt also said 'There are a lot of things about WISeR that are not working,' referring to CMS's AI-supported Medicare prior-authorization model, and the slideshow lists lessons: continuous oversight, essential human clinical review, and provider feedback as a monitoring signal. Why now: Indiana's pilot was announced June 24 as a 90-day effort, so its proof-of-concept report is due, and CMS says it intends to scale the model to more state Medicaid programs.
Why it matters
State Medicaid agencies own claims systems, payment holds and provider enrollment, so a federal prepayment AI platform would run on state data and change state workflows. Five additional states are already in pilots, and others may be asked to join. The WISeR admissions are the first official acknowledgment in this cycle that an AI-supported federal health program model has operating problems.
Evidence and measured results
Source findings (reported from Brandt's remarks and slideshow): Indiana pilot with Oracle complete or near-complete with results 'soon'; five other states in active pilots; AI flags suspicious billing before payment; shared investigation platform under test. Operator claims not attributable to the AI pilots: $41.9 billion FY2025 savings (59% increase) from the broader anti-fraud push; Medicaid War Room stopped $342 million in improper payments and sent 124 notices of intent to exclude as of September 2026; about 67% of watchlisted new providers and suppliers cut. Enrichment from Indiana FSSA's June 24 release: objectives include prepayment detection, recommended claim edits and prior-authorization triggers, AI-generated investigation-ready case packages, a scalability assessment and a final proof-of-concept report; WISH-TV (June 24) reported a 90-day term, removal of personal and protected health information before sharing with Oracle, and that the tool cannot initiate regulatory action. No accuracy, false-positive, baseline or sample data are available.
Limitations and uncertainty
Reporting of conference remarks and a slideshow; the slideshow was not inspected. No pilot outcomes, accuracy or false-positive data were released. The five other states are unnamed. Savings figures describe the broader anti-fraud program, not AI. Indiana pilot design comes from the June 24 state release and June WISH-TV reporting, not from results.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-18; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- State Medicaid agencies face pressure to adopt federal AI prepayment fraud detection without published evidence of its accuracy or a defined state review process.
- Stakeholders
- State Medicaid director
- Program integrity director
- Medicaid Fraud Control Unit
- State privacy officer
- MMIS vendor and state CIO
- Provider associations
- Discovery questions
- Has CMS approached the agency about joining a Medicaid AI pilot or platform?
- What evidence would the agency require before holding a claim on an AI flag?
- How are provider notice and appeal handled for prepayment holds today?
- What data-sharing terms govern claims data sent to federal or vendor analytics?
- Value hypothesis
- A short readiness review may help an agency set evidence, review and data-sharing conditions before joining a federal AI pilot; no savings are established by this source.
- Potential engagement
- Four-week Medicaid AI-flag readiness review: map claims-hold workflow, draft flag-review and appeal procedures, and list evidence and data-use terms to request from CMS.
- Unsupported claims
- That the Indiana pilot reduced fraud or improper payments
- That CMS's $41.9 billion savings came from AI
- Any accuracy or false-positive rate for the pilots
- That the five other states are named or committed to scaling
Pre-sales engineering
Role takeaway
- Fit
- Agencies whose MMIS supports prepayment edits and claim holds and that can exchange near-real-time claims data with external analytics.
- Architecture and integration
- Interpretation: claims feed to a federal or vendor analytics layer; flags return to MMIS prepayment edit queues; a shared case workspace links state program integrity, MFCU and federal users.
- Prerequisites
- Near-real-time claims extract from MMIS
- Configurable prepayment hold and release logic
- Data use agreement covering de-identification and destruction
- Role-based access model for shared case platform
- Constraints
- State Medicaid due-process rules for provider payment holds
- HIPAA minimum-necessary limits on shared data
- De-identified data must be re-linked before state action
- No published model performance data yet
- Security
- Reported Indiana terms remove personal and protected health information, restrict storage and bar onward sharing; a scaled platform with law-enforcement users needs access logging and least-privilege controls.
- Proof of value
- Compare AI flags with historical confirmed fraud cases to estimate precision
- Measure share of flagged claims overturned on human review
- Measure hold duration added to legitimate claims
- Test re-identification workflow for flagged de-identified claims
Delivery
Role takeaway
- Work
- Stand up a governed flag-review workflow: intake of AI flags, human review, hold or release decisions, provider notice, appeal and feedback to the model owner.
- Dependencies
- CMS pilot terms and data use agreement
- MMIS vendor change requests
- Program integrity staffing
- Legal review of payment-hold authority
- Ownership
- State Medicaid program integrity director, with the MMIS business owner and privacy officer as co-owners.
- Skills and adoption
- Train reviewers to treat flags as leads that need evidence, document reasons for each decision, and escalate patterns of wrongful flags.
- Governance checkpoints
- Review Indiana proof-of-concept results before joining
- Privacy and legal sign-off on data use terms
- Human review required before any prepayment hold
- Quarterly review of overturned flags and hold durations
- Acceptance criteria
- Every hold has a documented human reviewer and reason
- Proposed: median added hold time for released claims under a set number of days
- Proposed: overturned-flag rate reported monthly to leadership
- Provider appeal path published before go-live
- Risks
- Wrongful payment holds harm legitimate providers and access to care
- Results may not transfer from Indiana to other states
- Opaque federal models limit state ability to justify actions
- Pressure to scale before results are reviewed
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
A national prepayment platform implies near-real-time claims feeds from state MMIS systems to a federal or vendor-hosted analytics layer, with flags returned before adjudication. States will need defined interfaces for receiving flags, holding claims, and returning dispositions, plus a shared case workspace that spans state, MFCU and federal users with role-based access.
Governance
Who approves, reviews and stays accountable for outcomes?
Prepayment AI flags can delay payments to providers, so states need documented criteria, human review before holds, provider notice and appeal paths. CMS's own WISeR lessons (continuous oversight, human clinical review, provider feedback) are a reasonable minimum for any state joining a pilot. Pilot results should be reviewed for false positives before scaling.
Security and privacy
What data, permissions and controls need testing?
Indiana's data use agreement, as reported, de-identifies personal and protected health information, restricts storage and bars Oracle from sharing FSSA data. A scaled shared platform with law-enforcement users raises access-control, logging and minimum-necessary questions that each state's privacy office must review.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Indiana says the tool will not replace workers and cannot initiate regulatory action, so reviewer capacity becomes the constraint. Agencies should plan staff time to review flags and handle provider appeals, and keep provider notices in plain language.
Procurement
What should contracts, pricing and exit terms secure?
States may join federal pilots through data-sharing agreements rather than procurement, which can bypass usual vendor-evaluation steps. Agencies should require model documentation, performance reporting, data destruction and exit terms in any agreement.
Operating model
Which teams own the service once it runs?
A national platform would shift some fraud detection to CMS and vendors while states keep responsibility for payment actions. States need a clear split of who flags, who decides and who answers to providers.
Publication history
- 2026-10-18State Government · Issue 412 resources
Stable resource ID: cms-medicaid-ai-prepayment-fraud-pilots-indiana-results-pending-five-states-20261015