There is a particular species of silence that settles over a federal courtroom when a judge stops believing a prosecutor. It is not the silence of a technical objection or a routine sidebar. It is something colder — the sound of a presumption dying. For most of the modern history of American criminal law, the government's lawyer has entered the courtroom carrying an invisible endowment: the assumption, embedded in doctrine and habit alike, that when a federal prosecutor represents something to the court, it is true. Judges lean on that assumption because the alternative — auditing every representation, distrusting every filing — would grind the machinery of justice to a halt. But in the spring of 2025, according to reporting by Bloomberg Law, that assumption began to buckle in courtrooms across the country, and the sound it made was the sound of a system discovering that one of its oldest load-bearing beliefs had quietly rotted.
On or around May 21, a federal judge identified in the reporting as Judge Perry did something that, in another era, might have seemed almost heretical. Rather than extending the customary benefit of the doubt to the Department of Justice lawyers before the bench, the judge challenged the premise itself — the idea that federal prosecutors ought to be presumed trustworthy at all. It was a small gesture in the choreography of a single case, but it gestured at something enormous. If the presumption of prosecutorial good faith is not a fact of nature but a policy choice, then it can be revoked when the evidence warrants. And a growing body of evidence, chronicled in a Bloomberg Law analysis dated around May 27, 2025, suggests the warrant may be arriving.
That analysis examined what it described as a wave of misconduct by Justice Department attorneys — a wave made worse, in the reporting's telling, by the very deference that courts have long shown to prosecutors. The two phenomena are not separate. They are the same phenomenon seen from opposite ends. Deference is what makes misconduct cheap. And when misconduct becomes cheap, it multiplies. The question the reporting raises is not merely whether individual prosecutors have behaved badly. It is whether the entire architecture designed to catch them has been hollowed out at precisely the moment it is most needed.
The Endowment of Belief
To understand why a judge questioning a prosecutor's trustworthiness registers as a tremor rather than a routine ruling, you have to understand how much of the system runs on unexamined faith. The presumption of regularity — the doctrine that government officials, prosecutors chief among them, are assumed to have discharged their duties lawfully and in good faith — is not written into the Constitution in so many words. It is a working assumption, refined over generations, that allows courts to process an enormous volume of cases without treating every government filing as a potential fraud. In practice it means that when a prosecutor tells a judge that all exculpatory evidence has been turned over, the judge generally believes it. When a prosecutor represents that a witness has no undisclosed deal, the judge generally believes that too.
The genius of the arrangement is efficiency. Its vulnerability is obvious: the presumption is only as good as the honesty of the people it protects. A prosecutor who is willing to lie inherits, by default, the credibility that the system extends to prosecutors who are not. The dishonest borrow the good name of the honest. And because the presumption operates silently — no one announces that they are extending it — its abuse is silent too. There is no alarm that sounds when a false representation slips past a trusting judge. The misconduct simply disappears into the record, indistinguishable, for a while, from ordinary practice.
This is why Judge Perry's challenge matters beyond the four corners of any single docket. To question the presumption is to acknowledge, publicly and on the record, that the endowment of belief can be withdrawn — that trust is a thing a prosecutor's office can spend down. According to the reporting, the concern animating this skepticism is not abstract. It is the fear that malfeasance in politically sensitive cases is growing, and that the tools available to detect it are shrinking. A judge who says, in effect, I will no longer simply take the government's word is not grandstanding. That judge is responding to a structural reality: that a system built on faith is exquisitely vulnerable to those who exploit faith.
The Word for What They Did to the Court
There is a word that appears in the reporting, and it is a word chosen with care, because judges do not use it lightly. According to Bloomberg Law's analysis, at least one judge accused Department of Justice lawyers of gaslighting the court. It is worth pausing on the accusation, because it is not the language of ordinary judicial displeasure. Judges have a rich vocabulary for expressing dissatisfaction with the government — they can call a filing "unpersuasive," a position "not well taken," an argument "meritless." Those are the polite instruments of the bench. Gaslighting is not among them. It is a term borrowed from the study of psychological abuse, and it describes a specific, deliberate cruelty: the effort to make someone doubt their own perception of reality.
To accuse a lawyer of gaslighting the court is to allege something far graver than error. It is to say that the government did not merely get something wrong but attempted to manipulate the tribunal's grasp of the truth — to make the court distrust what it plainly saw. When a judge reaches for that word, the judge is describing conduct that attacks not just the outcome of a case but the epistemic foundation on which every case rests. A court that cannot trust the representations made to it cannot function as a court. It becomes a stage for performance rather than a forum for fact.
To accuse a lawyer of gaslighting the court is to allege not error but the deliberate manipulation of a tribunal's grasp of the truth — an assault on the very thing that makes a court a court.
The reporting frames this not as an isolated outburst but as a symptom. The core argument advanced in the analysis is deceptively simple and genuinely chilling: if the law makes it easy to conceal unethical or unlawful conduct by federal prosecutors, and if malfeasance in politically sensitive cases is on the rise, then the public almost certainly knows only a fraction of what is actually occurring. The cases in which a judge grows angry enough to use a word like gaslighting are, by this logic, not the iceberg. They are the tip of it — the rare instances where the misconduct was clumsy enough, or the judge attentive enough, or the record complete enough, for the concealment to fail. The successful concealments, by definition, leave no visible trace.
Brady's Broken Promise
The reason concealment is so consequential in the criminal context has a name, and the name is Brady. In Brady v. Maryland, decided in 1963, the Supreme Court held that the prosecution's suppression of evidence favorable to a defendant violates due process where the evidence is material to guilt or punishment. The principle is now so familiar that its radicalism can be forgotten. Brady imposes on the government an affirmative obligation not merely to refrain from lying but to hand over the very material that might undermine its own case. It asks the adversary to arm its opponent. It is one of the most idealistic commitments in American criminal procedure, and it depends almost entirely on the honesty of the party bound by it.
That is the flaw at the heart of the doctrine. A Brady violation is, in its nature, a crime of concealment. The defense often has no way of knowing that favorable evidence exists, because the only party who knows is the party withholding it. Discovery of the violation typically requires luck, a leak, a diligent post-conviction investigation, or a prosecutor's own belated attack of conscience. And even when a violation surfaces, the consequences for the offending lawyer have historically been mild to nonexistent. The remedy runs to the defendant — a new trial, a vacated conviction — but rarely to the prosecutor. The lawyer who withheld the evidence may face no professional discipline, no removal, no meaningful sanction at all. The reporting underscores this asymmetry: Brady obligations, it notes, are chronically under-punished.
The most instructive illustration in recent American memory is the federal prosecution of the late Alaska senator Ted Stevens. Stevens was convicted in 2008 on corruption-related charges shortly before he lost his Senate seat. It later emerged that federal prosecutors had failed to disclose exculpatory material to the defense — Brady violations that ultimately led the Justice Department itself to move to dismiss the case, and a federal judge to set the conviction aside. The Stevens affair became a byword for what can go wrong when the government's disclosure obligations are honored in the breach. It also demonstrated how much has to align for such misconduct to come to light: a high-profile defendant, aggressive defense counsel, a whistleblowing agent, and a judge willing to press. Strip away any one of those conditions and the violations might have remained buried, and a United States senator might have gone to prison on a tainted record.
The lesson the innocence movement has drawn from cases like these, over decades of documented exonerations, is that the Stevens prosecution was not an aberration in kind, only in visibility. Organizations that investigate wrongful convictions have repeatedly identified official misconduct, including the suppression of favorable evidence, as a recurring contributor to the imprisonment of the innocent. What distinguishes the cases we know about from the ones we don't is not the presence or absence of misconduct. It is the presence or absence of someone with the resources, the tenacity, and the good fortune to dig it up.
The Shield That Cannot Be Pierced
Suppose, though, that a defendant does uncover the misconduct — that the concealment fails, the evidence surfaces, the wrong is proven. Surely then the prosecutor who committed it can be held to account. Here the law offers one of its most confounding answers, and its name is Imbler v. Pachtman. In that 1976 decision, the Supreme Court held that prosecutors enjoy absolute immunity from civil liability for actions taken in their role as advocates — the initiation and pursuit of a criminal prosecution. Absolute immunity is exactly what it sounds like. It is not a defense to be weighed against the facts. It is a wall. A prosecutor who deliberately withholds exculpatory evidence, who knowingly presents false testimony, who pursues a case he knows to be baseless, is generally shielded from a civil suit for damages arising out of those advocacy functions, however egregious the conduct.
The Court's rationale in Imbler was not indifference to abuse. It reasoned that the threat of personal liability would make prosecutors timid, would flood the courts with retaliatory litigation from disgruntled defendants, and would deter vigorous law enforcement. The remedy for prosecutorial misconduct, the Court suggested, lay elsewhere — in the criminal process itself, in professional discipline, in the internal accountability structures of prosecutors' offices. In other words, Imbler did not deny that prosecutors could do terrible things. It relocated the responsibility for stopping them. It said, in effect: the civil courts will not police this, because other mechanisms will.
That relocation is the hinge on which this entire story turns. Absolute immunity is defensible only to the extent that the "other mechanisms" actually function. If prosecutors cannot be sued, then discipline, ethics review, and professional responsibility must do the work the civil courts are excused from doing. The Brennan Center for Justice, in its work documenting the weaknesses of Justice Department accountability, has long argued that these alternative mechanisms are thinner and more porous than the public assumes. The internal offices meant to catch and punish misconduct are under-resourced, opaque, and dependent on the goodwill of an institution investigating itself. The promise Imbler made — that other checks would fill the gap left by immunity — was always a promise on credit. And in the spring of 2025, according to the reporting, that credit came due.
Dismantling the Watchmen
The internal architecture of accountability inside the Justice Department is not a single office but a small constellation of them, each occupying a slightly different position in the constellation of oversight. There is an ethics function, charged with advising on and enforcing standards of conduct. There is a senior review layer, where the most sensitive and consequential judgments about attorney conduct and case handling are made. And there is the Office of Professional Responsibility, or OPR, the internal body specifically tasked with investigating allegations of misconduct by department lawyers. Together these offices constitute the machinery on which Imbler's promise depends — the "other mechanisms" that were supposed to justify absolute immunity.
According to the reporting, that machinery was subjected to a systematic dismantling. Joseph Tirrell, the department's Ethics Director, was removed. Bradley Weinsheimer, who occupied a senior ethics review role, was removed. And Jeffrey Ragsdale, who led the Office of Professional Responsibility, was removed as well. Read individually, each departure might be explained away as the ordinary turnover of a new administration reshaping its personnel. Read together, they describe something more troubling: the near-simultaneous removal of the very people whose job was to hold the department's own lawyers to account. When you remove the ethics director, the senior ethics reviewer, and the head of the office that investigates misconduct, you have not trimmed the bureaucracy. You have decapitated the internal check.
The significance of this is not merely that particular individuals lost their posts. It is what their removal signals about the incentives now facing every prosecutor in the department. Accountability structures do their most important work not through the cases they prosecute but through the cases they deter — through the quiet awareness, in the mind of a lawyer weighing whether to shade a representation or bury a document, that someone is watching and that consequences are possible. Remove the watchers, and you do not merely fail to punish the misconduct that occurs. You increase the misconduct that occurs, by lowering its expected cost toward zero. The removals described in the reporting are, in this sense, not the end of the story but the beginning of a new one, whose chapters have not yet been written into any visible record.
The Geometry of Invisible Abuse
What emerges from the reporting, when its threads are drawn together, is not a collection of scandals but a geometry — a set of interlocking conditions that, when they align, make prosecutorial abuse not merely possible but structurally invisible. Consider how the pieces fit. The presumption of regularity means courts extend the government the benefit of the doubt. Brady violations are, by nature, crimes of concealment that the victims often cannot detect. Absolute immunity under Imbler means that even proven misconduct rarely results in civil consequences for the prosecutor. And the internal accountability offices — the very mechanisms Imbler pointed to as the alternative remedy — have, according to the reporting, been stripped of their leadership.
Each of these conditions is troubling on its own. Any single one, in isolation, might be tolerable, because the others would compensate. If courts are too trusting, internal ethics review can catch what they miss. If immunity forecloses civil suits, professional discipline can still impose consequences. If Brady violations are hard to detect, a robust OPR can at least investigate the ones that surface. The system was designed with redundancy — with the understanding that no single check would be perfect and that the checks would cover for one another's failures. What the reporting describes is the collapse of the redundancy itself. It is the removal of every backstop at once.
This is the deeper argument beneath the individual facts. The danger is not any particular judge's anger, or any particular official's removal, or any particular Brady violation. The danger is the simultaneity. When courts' skepticism is dulled, and internal ethics offices are gutted, and professional responsibility is weakened, and civil immunity remains absolute, the failures do not merely add. They compound. Each weakened check removes the safety net for the others, until the remaining structure can no longer catch anything at all. Abuse, in such a system, does not need to be tolerated to flourish. It merely needs to be unobserved, and every mechanism designed to observe it has been switched off in the same season.
The reporting is careful, and so this account must be careful too. What is documented is a pattern of removals, a wave of misconduct allegations, judicial expressions of alarm as pointed as the charge of gaslighting, and a structural argument about how easily such misconduct can be hidden. What is not documented — what cannot be documented, given the very nature of the problem — is the full extent of what remains concealed. That is precisely the point Judge Perry's skepticism was reaching toward, and precisely the point the Bloomberg Law analysis drives home: in a system engineered to make misconduct invisible, the absence of proof is not the proof of absence. The cases that reach the light are the failures of concealment, not its scope.
What the Silence Means
Return, at the end, to that silence in the courtroom — the sound of a presumption dying. It is tempting to hear it as a loss, and in one sense it is. A justice system in which judges must approach the government's own lawyers with suspicion is a diminished thing, slower and more adversarial and more exhausting than the one the presumption of regularity was meant to enable. Every hour a court spends verifying what it once could assume is an hour not spent on the substance of justice. The erosion of trust is a real and heavy cost, and the prosecutors who spent down that trust — if the allegations in the reporting hold — have imposed it not only on themselves but on the honest majority of their colleagues, who will now be doubted for the sins of a few.
But there is another way to hear the silence, and it may be the more hopeful one. A judge who questions the presumption of prosecutorial good faith is not abandoning the system. That judge is defending it — refusing to let the endowment of belief be exploited, insisting that trust must be earned rather than assumed, treating the government as a party whose representations can be tested against the truth. In a landscape where the internal watchmen have been dismissed and the civil courthouse door is barred by immunity, the judiciary's willingness to withdraw its trust may be, for a time, the only check that remains standing. The word gaslighting, harsh as it is, is the sound of a court refusing to doubt its own eyes.
The founders of the American legal order understood that power concealed is power abused. They built a government of divided and overlapping authorities precisely because they distrusted the concentration of unaccountable discretion in any single set of hands. Prosecutorial power is among the most fearsome discretions the state possesses — the power to investigate, to charge, to shame, to imprison, and, through the strategic silence of a withheld document, to convict the innocent. That such a power should be hedged by every available check is not a partisan proposition. It is the oldest wisdom in the republic. What the reporting describes is the season in which the hedges came down. Whether they are restored, and how much abuse flourishes in the interval, will not be fully known for years, if ever — because the surest sign of a system that has lost its ability to see is that it can no longer tell you how much it has failed to see. The fraction we know is the measure of our luck, not the measure of the harm.