The lawyer became a conduit rather than a counselor
In Brazil, a formal complaint filed with the OAB by lawyer Schietti places a quiet but consequential question before the legal profession: when a machine authors the argument, who bears the responsibility? The petition at the center of the dispute carries no trace of human deliberation — only the assembled language of an algorithm — and in that absence, an entire framework of professional accountability is left suspended. What unfolds next may define not just Brazilian legal ethics, but the broader boundary between tool and replacement in the age of artificial intelligence.
- A lawyer has formally challenged the OAB over a petition written entirely by AI, with not a single sentence attributable to a human author.
- The complaint cuts deeper than procedure — it exposes a profession unprepared for the moment when its core intellectual labor is quietly outsourced to a machine.
- Brazil's bar association now faces an uncomfortable silence in its own rulebook: no clear standard yet exists for disclosure, review, or minimum human authorship in AI-assisted filings.
- The legal world is watching as AI drafting tools accelerate across the profession, blurring the line between assistance and wholesale replacement of human judgment.
- Schietti's decision to file formally rather than look away signals this is intended as a test case — one that will force the OAB to either act or, by its silence, permit the practice.
Brazilian lawyer Schietti has filed a formal complaint with the OAB — the country's bar association — over a legal petition that appears to have been generated entirely by artificial intelligence, with the submitting lawyer not having written a single sentence themselves. The complaint is not merely procedural. It asks what it means to practice law when the intellectual work at the center of that practice is handed over to an algorithm.
The petition in question bears the recognizable marks of AI generation: structured, formulaic, assembled rather than composed. Schietti's challenge goes to the heart of professional responsibility — a lawyer is expected to think through a case, weigh its nuances, and make deliberate choices about argument. When none of that occurs, the lawyer becomes a conduit rather than a counselor.
The OAB sets ethical standards and disciplines members who violate them, yet it has not yet fully answered the questions this case forces into the open: Must AI use be disclosed? Must documents be substantially revised by a human hand? Is wholesale AI authorship permissible at all? The association's response — or its silence — will itself constitute an answer.
The anxiety Schietti's complaint reflects is spreading globally. Legal AI tools are improving rapidly and are genuinely tempting for lawyers managing heavy caseloads. But there is a meaningful difference between AI as an aid to human judgment and AI as a substitute for it. Bar associations have long grounded their legitimacy in the idea that lawyers train, develop judgment, and bear responsibility for their work. A petition with no human fingerprints challenges that foundation entirely.
By filing formally rather than letting the matter pass, Schietti has made this a test case. How the OAB responds will shape the contours of professional responsibility in Brazil's legal world — and perhaps offer a signal to legal institutions everywhere navigating the same unsettled terrain.
A Brazilian lawyer named Schietti has filed a formal complaint with the OAB—the country's bar association—over a legal petition that appears to have been written entirely by artificial intelligence. According to Schietti's filing, the person who submitted the petition did not write a single sentence of it themselves. The complaint raises a fundamental question about what it means to practice law professionally when the work itself is outsourced to a machine.
The incident centers on a petition submitted to the OAB that bears all the hallmarks of AI generation: structured language, formulaic phrasing, the kind of document that reads as though it was assembled rather than composed. Schietti's challenge is not merely procedural—it cuts to the heart of professional responsibility. A lawyer is expected to think through a case, to understand its nuances, to make choices about argument and emphasis. When a petition is generated wholesale by an algorithm, none of that intellectual work has occurred. The lawyer becomes a conduit rather than a counselor.
What makes this complaint significant is its timing and its target. The OAB is Brazil's regulatory body for the legal profession. It sets standards, enforces ethics rules, and disciplines members who violate them. By filing this complaint, Schietti is essentially asking the association to confront a question it has not yet fully answered: what are the rules for AI in legal work? Is it permissible to submit an AI-generated petition at all? If so, under what conditions? Must a lawyer disclose that a document was machine-written? Must they at least review and substantially revise it?
The broader context matters here. Artificial intelligence is moving into professional spaces at remarkable speed. Legal research platforms are beginning to incorporate AI. Document drafting tools promise to speed up routine work. The technology is real, it is improving, and it is tempting—especially for lawyers managing heavy caseloads or tight budgets. But there is a difference between using AI as a tool to assist human judgment and using it as a replacement for human judgment entirely. Schietti's complaint suggests that line has been crossed in this case.
The complaint also reflects a professional anxiety that is spreading through the legal world globally. If a petition can be generated by AI without human authorship, what does that say about the value of legal expertise? What does it say about the profession's gatekeeping function? The OAB, like bar associations elsewhere, has built its legitimacy on the idea that lawyers undergo training, develop judgment, and bear responsibility for their work. An AI-generated petition with no human fingerprints on it challenges that entire framework.
Schietti's decision to report the matter formally rather than let it pass quietly suggests he sees this as a test case. The OAB will now have to respond. It may issue guidance on AI use. It may discipline the lawyer who submitted the petition. It may do nothing, which would itself be a statement. Whatever happens, the question is now in the open: in Brazil's legal profession, is there still a requirement that lawyers actually write their own work? The answer the OAB gives will shape how the profession adapts to artificial intelligence in the years ahead.
Citas Notables
Did not write a single sentence of it themselves— Schietti's characterization of the petition's authorship
La Conversación del Hearth Otra perspectiva de la historia
Why did Schietti decide to report this rather than just ignore it?
Because ignoring it normalizes the practice. If a petition can be submitted entirely AI-generated with no consequences, other lawyers will do the same. It becomes the new standard.
But couldn't a lawyer argue that AI is just a tool, like spell-check or legal research software?
There's a difference between using a tool to assist your thinking and replacing your thinking with a tool. A lawyer using AI to draft a first version and then substantially revising it is one thing. Submitting something the machine wrote without authoring a single sentence is another.
What does the OAB actually care about here? Is it about protecting lawyers' jobs, or is it about professional standards?
Both, probably. But the core issue is accountability. When you sign your name to a legal document, you're saying you stand behind it. You've thought it through. You're responsible for its accuracy and its arguments. An AI-generated petition breaks that chain.
Could this become a bigger problem? Like, could AI eventually be good enough that it doesn't matter?
That's the real question. If AI gets sophisticated enough, the distinction between "tool" and "replacement" blurs. But we're not there yet. And even if we were, there's a question about whether a profession should allow itself to be automated away entirely.
What happens next?
The OAB has to set rules. They might say AI is forbidden. They might say it's allowed only with disclosure and human review. They might do nothing and let the profession sort it out. But they can't ignore it now.