Skip to content
This is NOT an official site of the City of New York. Click here for official AI reports.

NYC Office of Technology and Innovation

NYC - DTPR AI Register

AI-Assisted Medicaid Lien Identification for Injury Cases

What it is for

Risk Assessment & Triage
  • MAVS scores individual Medicaid encounters as RELATED or UNRELATED to a personal injury for which a client received lawsuit funds, enabling DSS to prioritize lien placement decisions.

NYC Algorithmic Tools Compliance Report — MAVS

Eligibility & Public Benefits
  • The system's output informs lien decisions that determine whether a Medicaid client must repay benefits received for injury-related care, directly affecting the financial obligations of individuals who received Medicaid-funded treatment.

NYC Algorithmic Tools Compliance Report — MAVS

Run by
Department of Social Services (DSS)
Where
No fixed location
Kept
Retained Not specified in available disclosure
Shared with
Accountable organization
Your copy
You cannot see the data it holds about you. What you can do

What it is for

Risk Assessment & TriageEligibility & Public Benefits

This system uses machine learning to review Medicaid medical records and flag which healthcare encounters are related to a personal injury for which a client received lawsuit funds. The New York City Department of Social Services uses these results to determine whether to place a lien to recover Medicaid costs. A DSS staff member reviews every output before any notice is sent to a client or their attorney.

What it collects and what happens to it

Data taken in

Sensitive personal information
Identifiable data
  • Medicaid encounter data from the Medicaid data warehouse, including health records tied to specific individuals, and MAESTRO data containing personal injury case details, lawsuit settlement information, and prior lien determinations for identified Medicaid clients.

NYC Algorithmic Tools Compliance Report — MAVS

Sensitive personal information
Identifiable data
  • MAESTRO data on personal injury details including lawsuit settlement records tied to identified Medicaid clients, used alongside encounter data to match injuries to specific individuals.

NYC Algorithmic Tools Compliance Report — MAVS

Processing

Classification & Prediction
  • MAVS applies a machine learning classifier trained on historical MAESTRO determinations made by DSS staff. Each Medicaid encounter record is scored and assigned a binary label — RELATED or UNRELATED — indicating whether the encounter corresponds to the personal injury in question.

NYC Algorithmic Tools Compliance Report — MAVS

What it does

Deciding (Analytical AI)
Human decides
  • MAVS predicts and classifies each Medicaid encounter as RELATED or UNRELATED to a given personal injury claim. Under current practice, a DSS staff member reviews MAVS output before sending any notification, meaning the system is advisory and a human decides on downstream action.

NYC Algorithmic Tools Compliance Report — MAVS

Outputs

A decision about you
Identifiable data
  • Each Medicaid encounter in the client's MAESTRO file receives a determination of RELATED or UNRELATED. This output is reviewed by a DSS staff member and informs lien placement decisions that may require the client to repay Medicaid funds.

NYC Algorithmic Tools Compliance Report — MAVS

Run by

Department of Social Services (DSS)
  • The NYC Department of Social Services Accountability Office deploys and oversees MAVS to identify Medicaid encounters eligible for lien recovery, with staff review required before any client notification.

NYC Algorithmic Tools Compliance Report — MAVS

Built by

Gainwell / Health Management Systems
  • Gainwell/HMS supplies and maintains MAVS and the underlying MAESTRO system under a contract with the NYS Office of Medicaid Inspector General. DSS and OMIG hold regular meetings with Gainwell to discuss system performance.

NYC Algorithmic Tools Compliance Report — MAVS

Kept for

Retained Not specified in available disclosure
  • The register does not specify a retention period for MAVS output data. Data is noted as present in training data, input data, and output data, suggesting multi-stage retention, but exact periods are not disclosed.
  • Duration: Not specified in available disclosure

NYC Algorithmic Tools Compliance Report — MAVS

Shared with

Available to the accountable organization
  • MAVS output (RELATED/UNRELATED determinations per encounter) is available to DSS Accountability Office staff who review it before any downstream action. DSS and NYS OMIG also receive information on system performance through regular meetings with Gainwell.

NYC Algorithmic Tools Compliance Report — MAVS

Not available to me
  • Individual Medicaid clients do not have direct access to the MAVS score or RELATED/UNRELATED determination for their encounters. Clients are notified by DSS only after staff review, and only when a lien action is being initiated.

NYC Algorithmic Tools Compliance Report — MAVS

Stored

Not stated by the NYC Office of Technology and Innovation.

How to read the colours

Can it identify you?

Anonymized data
Data about people with the link to who is broken. Stripped of identifiers, blurred, aggregated, or noised so this system can’t reasonably tie a record back to an individual.
Pseudonymous data
Each person’s data is tied to a token (hash, ID, template) that lets this system recognise the same person across events, but the token itself doesn’t reveal a name. Reidentification is possible with extra information.
Identifiable data
The data either contains a direct identifier (name, address, account name, recognisable face or voice, plate number) or carries a token this system uses to look up legal identity during processing.

Who completes the loop?

Human decides
This mode suggests; a person decides what to do next. The AI is always advisory — a human is in the loop on every decision. Example: a triage tool ranks cases for a clinician who chooses which to see first.
Human executes
This mode decides; a person carries out the result. Example: an optimizer plans the day’s trash-collection routes, and drivers run them.
Autonomous
This mode decides and acts on its own. No person reviews each decision or carries out the resulting action.

Definitions from the DTPR standard. Amber is about your data, violet about who decides. The fuller the shape and the deeper the colour, the more identifying the data or the less a person is involved.

What you can do

Ask about this system

Questions go to the NYC Office of Technology and Innovation, not the vendor.

Your rights

  • Right to a Human ReviewUnder current practice, a DSS staff member reviews all MAVS output before any notification is sent to a Medicaid client or their attorneys. Clients who receive a lien notice may contact the NYC Department of Social Services to request review of the determination.
  • Right to Algorithmic TransparencyThe use of MAVS to identify Medicaid encounters related to personal injury claims is disclosed through the NYC Algorithmic Tools Compliance Report, publicly available at the NYC Open Data portal. Members of the public may review this disclosure to understand that machine learning is used in Medicaid lien identification.

Risks and safeguards

  • Reputational harmMisclassification of Medicaid encounters as RELATED may incorrectly flag clients for lien recovery, causing financial harm and reputational damage.Safeguard: Under current practice, a DSS staff member reviews MAVS output before sending any notification to the client or attorneys, providing a human check on each determination before consequences are triggered.
  • Financial & business harmIncorrect RELATED determinations could result in lien placement against Medicaid clients who are not legally obligated to repay, causing direct financial harm.Safeguard: The system is informative only — the Medicaid program's legal entitlement to lien is reviewed by DSS staff who verify MAVS output before initiating any recovery action or client notification.