Tuesday, six in the morning. The police knock at an executive’s door. They execute a search warrant. While the agents are still in the house, the wife’s phone starts receiving links: news portals are already publishing the operation, the name of the target, the name of the company, the internal code of the investigation, the amount allegedly moved, the nickname by which — according to the report — the “core of the scheme” called him. The children are woken by messages from college classmates. The lawyer, summoned in a rush, arrives at the property and discovers that the press already circulating knows more about the accusation than he does. The agents leave with the seized devices. The defense is left with the seizure in hand, without a copy of the case file, without formal detail of what motivated the operation — and having to respond, on the first journalist’s call, to accusations that the press received before it did.
The confidentiality of the investigation, in theory a constitutional guarantee of the accused (Article 5, LX, of the Brazilian Federal Constitution; Article 20 of the Brazilian Code of Criminal Procedure — CPP), existed for all practical purposes until the police arrived. From the moment the operation was launched, confidentiality began to exist in one direction only: for the target. Not for the readers of the newspapers who, that same morning, learned about the accusation with more detail than defense counsel had.
The repetition eliminates the accident hypothesis
An isolated leak is an accident. The repetition of the pattern over more than a decade — same sequence (launch of the operation followed by a detailed news report the same day, sometimes the day before), same level of detail (amounts, internal case codes, nicknames of the accused, contents of intercepted messages), same timing (the window in which the elements exist in the file but have not yet been formally communicated to the target or the defense), involving different authorities, in different jurisdictions, in investigations without any connection between them — is not accident. It is practice. What repeats with predictable regularity is not random; it is systemic. And what is systemic is not attributable to an isolated actor, but to a routine that reproduces itself because it produces results.
The point does not require accusing any specific police chief or prosecutor by name. The informal chain of access to inquiry information is long: press officers of investigative bodies, agents involved in executing warrants, support staff with system access, members of other bodies who received institutional sharing under Brazilian cooperation legislation, collaborators of the Public Prosecutor’s Office, forensic experts, third parties heard in the pre-procedural phase. Each link is a possible vector. The defense does not need to identify the specific link to recognize that the pattern exists.
What matters legally is not the author of the leak — it is the effect. And the effect is predictable enough to turn the leak, in practice, into a procedural tool. A tool that procedural law does not authorize, but that produces results which procedural law, if it authorized them, would produce by other means.
What the leak produces
Observing that the leak produces effects on the accused is trivial. What the defense needs to do is map, with precision, which effects — so it can respond to them. Four stand out.
Anticipation of social conviction. Publication of investigation details before defense counsel has access to the case file produces a public verdict without adversarial process. In the months or years separating the headline from the sentence, the accused is treated as guilty by clients, suppliers, credit committees, boards of companies in which he sits, audit courts, regulatory agencies. Acquittal, when it comes, does not reverse this perception. Media-and-law research documents the asymmetry: accusation goes to the front page; acquittal goes to the footnote, when it appears at all. Defense that ignores this works for a trial that, in public perception, has already occurred.
Pressure on plea-bargain cooperation. Article 4 of Brazilian Law No. 12,850/2013 (the Organized Crime Act) requires that cooperation be voluntary. Voluntariness is an exacting legal concept: it presupposes the absence of coercion, even indirect coercion, and the capacity to deliberate under conditions that preserve autonomy. An accused whose company has lost market value in forty-eight hours, whose partners have left, whose bank credit has been suspended, whose contracts have been terminated under morality clauses, does not deliberate under the same conditions in which he would deliberate without the leak. Cooperation begins to operate as a way out of suffering — not as a considered legal decision. The vice of voluntariness is born outside the case file, but it contaminates the procedural legal act. Defense that fails to document this context hands the prosecution negotiating conditions that the law does not authorize.
Contamination of the evidentiary horizon. Witnesses read the reports before being interviewed. Forensic experts absorb hypotheses of the journalistic narrative. Judges, including those at appellate levels, are also readers. The formal standard of impartiality survives, and this is not to accuse any specific judge of anything. What the psychology of judicial decision-making has documented for decades is that anchoring bias — the effect of initial information on subsequent judgments — operates below conscious control (Guthrie, Rachlinski, and Wistrich, Inside the Judicial Mind, 86 Cornell L. Rev. 777, 2001). Once the narrative frame is fixed by the news report, the evidentiary phase is read from within that frame. Defense that does not work to dismantle the initial anchoring works on ground already configured against it.
Irreversible economic damage. A company that loses market value in the forty-eight hours following the report does not recover that value with an acquittal three years later. Contracts terminated under morality clauses, financing canceled by risk committees, exit of strategic suppliers, cut off from public procurement, loss of public officials who can no longer interact with a company under publicly disclosed investigation — all of this occurs in the space between the headline and the first judicial ruling. In that space, there is no judge to decide and no due process to invoke. There is only damage. Civil liability for the leak is potential, future, and limited reparation; it is not a procedural remedy.
The silence the defense must break
The confidentiality of the investigation is inscribed in Article 5, LX, of the Brazilian Federal Constitution, Article 20 of the CPP, and Article 7, § 10, of Brazilian Law No. 8,906/94 (the Statute of the Bar), which secures counsel’s access to the elements already documented. The breach of confidentiality by a public official is a functional crime under Article 325 of the Brazilian Penal Code, and an administrative infraction under the statute of each civil-service career. In the formal plane, the protective framework exists.
In the practical plane, this framework is almost never activated. Habeas corpus against specific acts, yes; defensive precautionary measures to preserve reputation while the investigation runs under formal confidentiality, almost never; criminal complaint for breach of functional confidentiality with an effective request for investigation of the origin, extremely rare; autonomous theory of nullity of the prosecution based on contamination arising from the leak, thin and resisted. The Brazilian Federal Supreme Court (STF) and the Superior Court of Justice (STJ), when they confront the matter, have been conservative in recognizing the leak as an autonomous cause of nullity. They recognize the functional crime; they separate it from the process. The defense has tacitly accepted this separation.
The thesis that Brazilian corporate criminal defense needs to begin constructing is that systematic leaking violates due process of law in its substantive dimension — because it distorts the material conditions under which the defense is exercised. Article 5, LIV, of the Brazilian Federal Constitution guarantees due process. Due process is not just the rite. It is also the set of conditions that make the rite capable of producing a just decision. When those conditions are deliberately distorted before the formal beginning of the process, the guarantee is violated in its material dimension — even if preserved in its procedural dimension.
This thesis is not consolidated. It needs to be articulated in each concrete case, with documentary proof of the pattern and the specific effect produced. It is not a textbook thesis. It is frontier thesis, and like every frontier thesis, requires systematic invocation before it becomes settled case law.
What the defense does when it finds out about the operation from the newspaper
The correct framing of the problem changes the defense’s script. What is described below is what corporate criminal defense operating in the real scenario — in which the press knows before the defense — actually executes. These are not recommendations. They are procedures that integrate the defensive act itself, from hour one.
Immediate collection of the leaked narrative as evidentiary material. All reports published in the critical window — from the eve of the operation to the end of the second day — are preserved in full copy with record of date, exact time, outlet, authorship, and comparison against the known procedural chronology. This material is a technical piece of the defense, not a journalistic chronicle. It serves to demonstrate, in court, three independent facts: that the information existed in the file before it existed for the defense; that the level of detail is incompatible with any hypothesis other than access to the file; and that the temporal window of disclosure coincides with the known procedural movement. Without this immediate collection, the leak dilutes and becomes “press coverage” — a legally inert category.
Public response coordinated technically by the defense. The accused’s institutional communication — statement, interview, clarification — passes entirely through defense review before going public. Not because the defense does communication, but because every word that comes out in that moment can be read against the accused in court. This means: no statement confirms hypotheses of the leaked narrative; no statement produces usable admission; the communication distinguishes with clarity what is asserted as fact from what is sustained as legal reading; and silence is a legitimate communicational decision when speaking produces more risk than gain. The typical error of hour one — press office speaking before the defense understands the case — creates material for the prosecution to process later.
Immediate request for access to the case file. Defense files, at hour one, a substantiated motion for full access to the case file based on Article 7, § 10, of Law No. 8,906/94 and Binding Precedent No. 14 (Súmula Vinculante 14) of the STF. If access is denied or restricted beyond what is reasonable, the defense files a constitutional complaint (reclamação). This is not a ceremonial step — it is the acquisition of the minimum parity that the leak has subtracted. As long as the defense does not have the file, the prosecution has already had time to selectively leak to the press; while that lasts, the asymmetry operates entirely against the accused.
Criminal complaint and institutional communication for the breach of confidentiality. The material collected in the first step instructs, still at hour one, a criminal complaint for breach of functional confidentiality (Article 325 of the Penal Code) directed to the competent authority — normally the Prosecutor General, the internal-affairs unit of the body to which the official belongs, or the natural court. Simultaneously, institutional communication to the Brazilian Bar Association (OAB), to the state institutes of lawyers, to the criminal-defense associations. The objective is not only individual accountability (which rarely occurs) — it is the documentary production that defense formalized its dissent, and that this dissent enters the procedural record. Absence of that invocation is itself prejudicial: silence of the defense in the face of a leak is often read as tacit acquiescence.
Repositioning the calendar — treating hour one as prepared. In the real scenario, defense that finds out from the newspaper has arrived late. Defense that was in place before — because the company operates in an exposed sector, because compliance had signaled risk, because correlated investigations against competitors had already been launched — arrives on time. This is not paranoia; it is corporate hygiene in an environment of systemic investigation and systemic leaking. Exposed companies without previously constituted criminal defense treat hour one as an emergency. Exposed companies with defense already in place treat hour one as an already-mapped procedure.
The point that runs through these five items is one: defense that treats the leak as part of the process, and not as press coverage, administers the prosecution from the moment it becomes public. Defense that treats the headline as news loses at hour one.
A note on guarantees that articulate with each other
The article published in this blog last week addressed another formal guarantee whose systematic defensive mobilization is under construction — the inviolability of the domicile under Article 5, XI, of the Constitution, in the context of entry without a warrant and vitiated consent. The relationship between the two articles is not thematic: it is structural. In both, there is a consolidated constitutional guarantee; there is an equally consolidated pattern of practical erosion; and there is a deficit of defensive mobilization in the technical articulation that would hold the erosion accountable. The corporate criminal defense that occupies both fronts — illegitimate entry and systematic leaking — works on the frontier where formal guarantees need to be recovered as real guarantees. That article, in Portuguese, mapped the moment when the authority arrives at the door. This one maps the moment when the press arrived before.
The confidentiality of the investigation is a constitutional guarantee of the accused. It is not a procedural privilege of the prosecution, and it is not disposable property of law enforcement. When it is broken selectively to produce effects that procedural law would not authorize by other means, criminal procedure is being instrumentalized before it formally exists for the defense. Corporate criminal defense that recognizes this — that documents, invokes, coordinates communication and technique, and repositions its calendar to begin before the headline — operates at a structural advantage. Defense that continues to treat the headline as news loses before the first judicial order.
Guilherme Brenner Lucchesi is a criminal defense attorney, holds a doctorate in Law from UFPR, an LL.M. from Cornell Law School, is a professor of Criminal Procedure at the Federal University of Paraná (UFPR), president of the Paraná Institute of Lawyers (IAP), and founding partner of Lucchesi Advocacia.
