Mata v. Avianca Sanctions: What a Verification Certificate Would Have Changed
This piece is a thought experiment. Mata v. Avianca is a real case. The brief was real. The order to show cause was real. The sanction was real. The Verification Certificate, as described in this series, did not exist at the time the brief was filed. What follows is a reconstruction of the proceeding under one counterfactual change: the firm files a contemporaneous Verification Certificate alongside the brief.
The point of the exercise is not to relitigate Mata v. Avianca. The point is to show what the artifact would have changed and what it would not have changed. The certificate is not a shield. It is a record. The two are easy to conflate and worth keeping apart.
The factual backbone of this piece comes from the public record of the case, summarized previously in Anatomy of a Hallucination. The artifact described here is the same one described in What a Verification Certificate Contains. The verification step itself is the one described in The Pre-Filing Verification Step That Catches What Westlaw Does Not. The doctrinal framing is the one set out in Rule 11 and the Generative-AI Duty of Inquiry.
The counterfactual frame
The actual record in Mata v. Avianca contains a brief and an order to show cause. The brief cited eight cases. Six of the eight could not be located in the reporters the brief named. When the court ordered counsel to produce the opinions, counsel produced typed excerpts that did not appear in any reporter. The eventual acknowledgment was that the cases had come from a generative model during drafting and had not been verified against the reporter before signing.
The counterfactual change is small and singular. Assume the firm ran a pre-filing citation check on the final draft. Assume the check produced a Verification Certificate. Assume the certificate was filed alongside the brief in the same docket entry, or retained in the matter file and producible on order. Nothing else about the matter changes. The same Montreal Convention argument is made. The same attorneys sign. The same draft is filed.
The question is what the certificate would have shown, and what the proceeding would have looked like once the certificate was on the record.
What the certificate would have shown on the actual draft
The certificate, as a document, would list each citation extracted from the brief. Each row would carry a verdict. The verdicts use the hedged register described earlier in the series. A citation is located in the reporter at the cited volume and page, or it is not. The cited language matches the language in the opinion, or it differs, or no opinion is available against which to compare.
On the actual Mata draft, the certificate would have returned a verdict of citation not located in reporter on six rows. The other rows would have resolved cleanly. The certificate would have carried a SHA-256 hash of its content, a public token URL, and a timestamp generated at the time of the check. The recommendation column on the six unresolved rows would have read in the same direct register the certificate uses elsewhere: confirm in the reporter before filing.
That is the artifact. The artifact would have existed before the brief was signed. The associate would have read it. The partner would have read it. The act of signing would have occurred in full view of the record.
Scenario A: clean verdicts on all citations
Begin with the version of the counterfactual most favorable to the firm. Assume the certificate, somehow, returned clean verdicts on all eight citations. Each row resolved. Each citation matched its reporter. The certificate was filed with the brief.
In this version, the predicate for the order to show cause largely disappears. The court's inquiry in the actual case began when opposing counsel could not locate the cited opinions and flagged the discrepancy. In the counterfactual, the firm has on the record a contemporaneous artifact that the citations were checked and that the check returned clean. Opposing counsel might still raise an objection. The court might still ask questions. But the conversation is different from the one that occurred.
The court is not asking the firm to explain an absence. The court is asking the firm to reconcile a discrepancy between the certificate's verdict and a later finding. That is a narrower inquiry, and the firm has a starting point that the actual record did not contain: a dated, hashed record of a verification step that returned a result.
Sanctions in this version are unlikely. The duty of inquiry was discharged. The artifact records the inquiry. The error, if there is one, would be in the verification mechanism, not in the firm's process. That is not the fact pattern that produced sanctions in the actual case.
This scenario is also the least realistic. The citations in Mata v. Avianca did not exist in the reporters named. A scanner that resolves citations against reporter sources does not return clean verdicts on citations that are not in the reporter. The purpose of laying out Scenario A is to show what a clean record buys the firm, not to suggest that this draft could have produced one.
Scenario B: unresolved verdicts and the firm files anyway
The realistic counterfactual is the harder one. The certificate ran on the actual draft. The certificate returned six unresolved verdicts. The firm filed the brief anyway.
The order to show cause in this version is about a different question. The actual order was, in substance, asking how the attorneys came to sign a brief whose citations did not exist. The implied question was whether any meaningful verification step had occurred. The counterfactual order is asking why the firm filed a brief whose citations had been flagged as unresolved by its own pre-filing check. The implied question is no longer about the absence of process. The implied question is about a deliberate decision in the face of an adverse process output.
The firm's posture is worse in some ways and better in others. It is worse because the decision to file was deliberate. The certificate told the firm what was wrong. The firm filed anyway. The court is no longer dealing with negligence or inadvertence. The court is dealing with a choice.
The firm's posture is better in the narrow sense that the record is transparent. The firm did the check. The check produced a record. The record is on the docket or in the matter file. The firm is not in the position the Mata attorneys actually occupied, where the record had to be reconstructed under questioning and the explanation depended on what the attorneys remembered about the drafting process.
The likely sanction in this version is narrower than the actual sanction, and it lands differently. The actual Mata sanctions were premised in part on a failure to investigate the citations after opposing counsel raised the issue. The counterfactual record removes the question of what the firm knew and when. The firm knew at the time of filing. The certificate is dated. The verdicts were on the record. A sanction in this scenario is narrower because it does not need to make findings about process. A sanction in this scenario is also harder to contest, because the predicate fact is the certificate itself, not a reconstruction of the firm's internal process.
What the counterfactual teaches
The lesson of the two scenarios is not that the certificate prevents sanctions. The certificate does not prevent anything. The certificate records what the verification step found. The decision to file or withhold the brief still belongs to the firm. The duty of inquiry under Rule 11 still belongs to the signer. The certificate is the document that lets a court see what was actually done at the time the duty was performed.
Both counterfactual outcomes are better than the actual Mata outcome. That is the part worth holding onto. The actual outcome was the absence of any contemporaneous record of inquiry. The court was left to reconstruct, under questioning, what the firm's process had been. The reconstruction was not favorable to the firm. The published order is, in many ways, a record of the reconstruction itself.
Scenario A converts the proceeding into a discrepancy inquiry. Scenario B converts the proceeding into a narrower question about a deliberate filing decision. Both are improvements on a proceeding in which the firm cannot produce any record of the inquiry it owed.
That is what the artifact does. It moves the conversation from the question was the inquiry performed to the question what did the inquiry find. The first question is existential for the firm. The second question is technical. Firms can answer technical questions. Firms have a harder time answering existential ones under oath.
A closing note on the series
This is the eighth and final post in the series. The series began with Mata v. Avianca because the case is the canonical example. The series has worked through the doctrinal frame around Rule 11, the operational mechanism of a pre-filing citation check, the structure of the Verification Certificate, the taxonomy of verdicts the certificate uses, the standing orders from federal judges that have shaped the institutional posture, and the sanctions cases that have followed Mata. The series ends here with Mata again, viewed under a counterfactual.
The throughline is short. The audit layer is the operational response to a failure mode that the profession has now observed enough times to name. The certificate is the artifact the audit layer produces. Veritas runs both. The product does not write the brief, format the citations, or argue the position. The product runs the verification and produces the record.
The record is what the court reads when the question is asked later. The firms that produce the record do not need the counterfactual. The record is already there.