Brazil Labor Court Fines a Lawyer for AI-Faked Cites | TLY

AI Regulation Tracker  /  Court ruling

Brazil's top labor court fines a lawyer and company for AI-invented citations

The Tribunal Superior do Trabalho penalized a telecom and its attorney for citing precedents that did not exist, then sent the matter to the bar and federal prosecutors. The verification duty falls on the lawyer, whether or not AI was used.

Brazil's top labor court fines a lawyer and company for AI-invented citations regulation briefing
The Leveraged Years AI Regulation Tracker

Brazil's highest labor court has fined a company and its lawyer for building an argument on court decisions that were never handed down. The Tribunal Superior do Trabalho, ruling in its Sixth Panel, found that a telecom firm and its attorney cited precedents that do not exist in the court's jurisprudence system, and it treated that conduct as bad-faith litigation rather than an innocent slip.

What the court decided

In Processo RR 0000284-92.2024.5.06.0351, decided around March 10, 2026, the panel imposed a fine of 1 percent of the case value on both the company and its attorney for litigância de má-fé, the Brazilian doctrine that penalizes bad-faith use of the courts. The significance is in who was punished. The client did not escape because a lawyer drafted the filing, and the lawyer did not escape by pointing at a tool. Both bore the penalty.

The court also looked past the fine. It ordered ofícios, formal notices, to the Ordem dos Advogados do Brasil, the national bar, and to the Ministério Público Federal, the federal prosecution service, so that each could examine whether disciplinary or criminal rules were broken. A monetary sanction in one case has become the trigger for two separate investigations, one professional and one potentially criminal. That escalation is the part practitioners should notice most. A bad-faith fine is a cost of doing business that some litigants might absorb. A bar inquiry that can reach suspension, and a prosecutorial file that can reach criminal charges, are a different order of risk, and they now flow from the same set of fabricated citations.

Why the panel called it deliberate

What moved this from carelessness to bad faith was the internal evidence in the citations themselves. According to the court and reporting by Migalhas, one fabricated precedent was attributed to a judge and dated after that judge had already retired. A ruling cannot be authored by a retired member on a date when the member no longer sits. The panel read that impossibility as a sign of invented content created to influence the judgment, not a citation pulled from the wrong file.

That framing matters for every professional who uses generative tools. The defense that "the software made it up" did not reduce responsibility. The panel held that the duty to confirm the accuracy of what is filed rests entirely with the lawyer and the party, whether or not artificial intelligence was involved. The obligation is non-delegable. It attaches to the signature on the brief, not to the method used to write it.

What the ruling does not do

The decision does not ban the use of AI in legal work, and it does not create a new rule specific to AI. It applies an existing bad-faith standard to a new fact pattern. That is the point worth absorbing. Lawyers who assumed that AI misuse sits in a regulatory gray zone now have a concrete data point: existing litigation-conduct rules already reach fabricated citations, and a superior court is willing to apply them at full force, including referrals that carry professional and criminal weight.

It also does not require any particular verification technology. The court did not prescribe a tool or a checklist. It described a result the profession must reach, which is that every cited authority must actually exist and say what the filing claims it says. In practice, that means treating any AI-produced case list as a lead to be checked rather than a source to be quoted. Pull the decision from the issuing court's own database, confirm the parties, the date, and the holding, and keep a record that the check was done. The retired-judge citation in this case would have collapsed on the first such lookup.

The cross-border read for US professionals

US lawyers will recognize the shape of this case. It echoes Mata v. Avianca, the 2023 New York federal matter in which attorneys were sanctioned for submitting AI-generated cases that did not exist. Brazil's TST reaches the same conclusion but pushes further than most US Rule 11 sanctions by pairing the fine with referrals to the bar and to federal prosecutors. For a US firm with Brazilian operations, or one relying on local counsel there, the practical signal is that the consequences for fabricated citations in Brazil can extend beyond a fine into disciplinary and criminal channels. The verification duty is the same on both sides of the border. The downside in Brazil is broader.

Frequently Asked Questions

What exactly did Brazil's Superior Labor Court decide?

The Tribunal Superior do Trabalho, Sixth Panel, fined both a telecom company and its attorney 1 percent of the case value for bad-faith litigation after they cited court precedents that do not exist, and it referred the lawyer to the national bar (OAB) and federal prosecutors (MPF) for investigation.

Who is affected by this ruling?

Every lawyer and party filing in a Brazilian court. The court held that the duty to verify cited authorities is non-delegable and rests on the subscribing attorney and the client, regardless of whether an AI tool was used to prepare the filing.

Does using AI reduce a lawyer's responsibility for a fabricated citation?

No. The panel found that responsibility for confirming the accuracy of filed information stays entirely with the lawyer and the party. Blaming the software did not lessen the sanction.

How does this compare to what US courts have done?

It parallels US sanctions such as Mata v. Avianca, where attorneys were penalized for AI-invented cases, but Brazil's court went further by adding referrals to the bar and to federal prosecutors, creating disciplinary and potential criminal exposure on top of the fine.

Browse the full AI Regulation News tracker

Informational analysis for working professionals, not legal advice. Confirm how any rule applies to your situation with qualified counsel.