WOLF AI field guide
Healthcare and Medicaid fraud evidence workflow.
A healthcare or Medicaid fraud evidence workflow organizes clinical, billing, policy, and communications records so a legal team can compare claimed conduct with source material and investigate anomalies.
Published · Last reviewed
Written and maintained by the WOLF AI product team and checked against the public sources cited on this page. It has not been reviewed by outside counsel, and it is not legal advice.
What do the numbers say?
Of the more than $6.8 billion in False Claims Act settlements and judgments the Department of Justice reported for fiscal year 2025, over $5.7 billion related to matters involving the health care industry, restoring funds to programs including Medicare, Medicaid, and TRICARE.
Source: DOJ: False Claims Act settlements and judgments exceed $6.8B in fiscal year 2025
According to the Department of Justice, False Claims Act settlements and judgments exceeded $6.8 billion in the fiscal year ending September 30, 2025 — the highest single-year total in the history of the statute. Settlements and judgments since the 1986 amendments now exceed $85 billion.
Source: DOJ: False Claims Act settlements and judgments exceed $6.8B in fiscal year 2025
Alongside the 1,297 whistleblower filings in fiscal year 2025, the Department of Justice reports that the government opened 401 new False Claims Act investigations, a volume that sets the pace plaintiff-side teams are working against when a matter is under seal.
Source: DOJ: False Claims Act settlements and judgments exceed $6.8B in fiscal year 2025
Who is this workflow for?
Firms handling healthcare and Medicaid FCA matters, where the record set mixes clinical documentation, billing data, program policy, and internal communications, and where the comparison across those four is the whole case.
When it is the wrong tool. It is not a coding audit, a clinical opinion, or a substitute for a qualified coding or medical expert. It is also not the right shape for very large structured billing datasets, which need a separately scoped ingestion approach.
What documents and inputs do you need?
- Clinical records and notes, as PDFs, including scanned documents handled with OCR.
- Billing and claims exports converted to PDF, with the field meanings the team is working from.
- Program policy and payer requirements relevant to the period.
- Internal communications: compliance emails, memos, and escalation threads.
How does the workflow actually run?
- 1.Define the matter scope, custodians, time period, and supported document set.
- 2.Organize billing references, clinical records, communications, and policy documents.
- 3.Flag candidate anomalies and link them to the exact supporting passages.
- 4.Have counsel and qualified subject-matter experts validate significance and pursue missing evidence.
What does this look like in practice?
Illustrative scenario
Illustrative only. A Medicaid matter covering eighteen months of claims for one service line.
- Claims are grouped by provider and date and set against the clinical documentation for the same day.
- Candidate anomalies surface: claims on dates with no corresponding note, and a service code used at a rate that stands out within this record set.
- Each anomaly links to the claim line, the missing-or-present note, and the policy paragraph defining the code.
- A coding expert reviews the sample and rules out two of the three patterns as documentation practice rather than billing conduct.
What this does not show. An outlier within one record set is a question to investigate. It is not a benchmark, a statistical finding, or evidence of fraud.
What do you get out, and who reviews it?
- An organized record set aligned across clinical, billing, policy, and communications material.
- A list of candidate anomalies, each linked to the passages behind it.
- Counsel and qualified experts own significance, methodology, and any clinical or coding conclusion.
Where does this approach break down?
- An anomaly is not proof of fraud, falsity, knowledge, or damages.
- WOLF AI does not provide medical, coding, or legal opinions.
- Large structured billing datasets may require a separately scoped ingestion workflow.
- Protected health information handling must be agreed before any records are shared; see the security page for current controls and what is not yet certified.
Common questions
Is WOLF AI HIPAA certified?
No. There is no HIPAA attestation and no completed SOC 2 audit. The security page states the current controls plainly and lists what is not claimed.
Can it read a billing CSV directly?
Not today. Exports are converted to PDF first. Native CSV and DOCX ingestion is on the roadmap and is not represented as shipped.
Does an outlier mean upcoding?
No. It means a pattern in this record set differs from the rest of this record set. Whether it reflects billing conduct, documentation practice, or case mix is for qualified experts and counsel to determine.