FOR QUI TAM COUNSEL

Verify the relator's account against the public record, before you sign the complaint.

A qui tam case is built on two things you did not generate yourself and cannot fully verify by reading: the relator's account of the fraud, and the regulatory framework that makes it actionable. You plead both with particularity, and you sign the result.

Rule 9(b) forces specifics into the complaint: the false claims, the dates, the amounts, the billing codes, the actors, the regulatory hooks. A single fabricated or misremembered particular becomes a Rule 11 representation you made to the court. The defendant moves to dismiss on exactly that particular, and the work behind it was contingent. AI now drafts the complaint, summarizes the relator's documents, and assembles the regulatory scaffolding, producing fluent, confident specifics whose errors are invisible on a read. Checking the draft with another AI does not solve this: a second model trained on overlapping data shares the same blind spots and confirms the error with equal confidence.

The AI is the witness. The math is the judge. GauntletScore verifies each claim against primary sources, then computes the trust score with deterministic math. The agents gather the evidence; no model sits in the verdict. The same evidence always produces the same score.

What GauntletScore checks
Regulatory and statutory references against current regulatory text, so a draft does not cite a rule that has changed or quote one that does not say what it is claimed to say.
Case citations against court records, so a supporting authority resolves to a real decision that holds what the draft asserts.
Named entities and actors against the public record, so a provider, corporate parent, or excluded individual is the entity the complaint says it is.
The dates and sequence the fraud theory depends on because an FCA theory often turns on what happened before what.

GauntletScore runs a dedicated analysis of cause-and-effect claims, testing whether the timing, proportionality, and logic actually hold. A broken causal chain counts heavily against the document. The engine earns its value by catching the argument that reads well and does not hold, which is the argument that gets a complaint dismissed.

What the first run looks like

Upload the document at gauntletscore.com. Minutes later you are reading the Gauntlet Report: a trust score with a 95% interval, every flagged claim with its verdict and the primary source behind it, and a cryptographically signed, tamper-evident certificate. You sign knowing every particular was checked, or you catch the one that was not before the defendant does, and the certificate is available to you, to co-counsel, and to the government as it weighs the case. In a live production run, the engine examined nineteen claims in one fluent, credible document and returned two debunked against the court record, each with the source that contradicts it. Every voting agent independently recommended against proceeding.

LIMITS

It does not assess the merits, it does not replace your judgment about the theory, and it does not verify what no public source can confirm. A claim it cannot check is reported as unverifiable, not as false.

Where it fits
Intake.

Run the checkable specifics in the relator's account against the public record before committing the firm's contingent time.

Pre-filing review.

Verify every citation, regulatory reference, and pleaded particular before the complaint is filed under seal and signed.

Government presentation.

Hand over a signed certificate showing the specifics were independently checked, alongside the disclosure statement.

There is a draft complaint or a relator memo on your desk right now with particulars you have not tested. Run it. Either they hold and you sign knowing it, or one does not, and you found it before the defendant did.

Analyze a Document
Three free credits. No card. No call.