Brady Violations Are Structural, Not Rogue

The hardest part of litigating a Brady violation isn't the doctrine. It's the fact that the evidence you needed never reached the prosecutor who looked you in the eye and said "that's everything we have." The data now confirms what defenders have long suspected: this isn't a few bad apples. It's the orchard.

Professor Jennifer Mason McAward's forthcoming Vanderbilt Law Review study assembled the largest Brady database ever built: 386 adjudicated violations between 2004 and 2022. The Death Penalty Information Center has separately identified more than 200 death-sentenced people whose convictions or sentences were overturned for Brady-related misconduct. The scale is the point. A violation this frequent follows a pattern, and the pattern points at structure, not villainy.

Three numbers from that database should change how you prepare every case:

• 49% of violations occurred in homicide cases, the docket where discovery volumes are largest and stakes highest.
• Law enforcement, not the prosecutor, was the suppressor in 31% of violations, and where police acted in good faith, inadequate training and resources were at fault 43% of the time.
• In 51% of successful Brady claims, the withheld material never even reached the line prosecutor.

That last number is the one to sit with. Half the time, the ADA you're negotiating with genuinely doesn't know what exists. Professor Adam Gershowitz describes the system as "a jigsaw puzzle with a thousand tiny pieces" where "no one is really in charge." McAward calls prosecutorial teams "just a collection of parts that do not work well together." The structural drivers are concrete: Cook County prosecutors carry 300 or more open cases, Fort Worth misdemeanor prosecutors juggle 1,200 to 1,500 matters, large DA offices run 15 to 33% vacancy rates, and the Los Angeles prosecutor's office coordinates with nearly 100 separate law enforcement agencies.

There's also a documented cognitive-bias component. McAward observes that prosecutors charge defendants only when convinced of guilt, so it's "very hard to appreciate the importance of evidence that runs contrary to the theory of the defendant's guilt." An officer's report that contradicts the bodycam doesn't read as Brady material to someone who's already certain. It reads as noise.

What does this mean for your practice? If the failure mode is systemic, the response has to be systemic too. Waiting for the state's production and trusting it covers a workflow that assumes the line prosecutor is the only disclosure risk, when the data says police files are the gap half the time. That's why we built our Brady list research workflow around cross-checking what you actually received: comparing the police report against the bodycam, the 911 audio against the narrative summary, flagging inconsistencies before anyone asks. The violations are baked into caseloads and coordination gaps. Your review process has to assume they're coming, and go find them.

The Jencks-vs-Brady Timing Split Is Your Lever

A live circuit split gives you a concrete timing argument in 2026, and most defenders never press it. The question is whether Jencks Act timing or Brady timing controls when the government must hand over exonerating evidence wrapped in a witness statement.

The answer depends entirely on where you practice. Practitioner analysis of the split maps it cleanly: the Fifth, Sixth, and Ninth Circuits hold that Jencks trumps Brady, while the Second, Third, and D.C. Circuits hold that Brady trumps Jencks. Six circuits (First, Fourth, Seventh, Eighth, Tenth, Eleventh) haven't ruled at all.

That gap is the lever. Here's how prosecutors use it against you:

• They wrap core Brady material (exonerating facts) inside interview memoranda and grand jury transcripts, then invoke the Jencks Act timeline: production only "until after the witness has testified on direct examination."
• They conflate impeachment material with core Brady, blurring the line so a delayed disclosure looks like a judgment call rather than a constitutional violation.
• They offer "viewer" basis production on platforms like USAfx: you can look, but you can't download or print. You dictate or take notes, and the clock keeps running.

The Bradley authors call this maneuver "a façade of compliance with the effective use standard," and they're right. It also inoculates the government from later discovery-violation claims, because the production log looks clean.

Your 2026 move is to preserve the timing objection at the outset. Cite adverse circuit precedent by name in your first discovery motion. Force the court to distinguish core Brady from impeachment material on the record. If you're in a no-ruling circuit, you're building the appellate issue in real time.

Track production method, too. Log the date, the access limitations, the items you could review but not keep. The government's obligation is constitutional, not statutory, and defense practitioners argue it's inconceivable that a statutory timeline should override a due-process one.

On the review side, tag every withheld item as core Brady or impeachment and note the required disclosure window. A Brady list research workflow that logs production method, date received, and access limits builds the record you need to argue "effective use" was denied. Evidence Copilot's auto-signals on ingest also surface inconsistencies between a police report and bodycam footage before you ask, which supports the Kyles collective-consideration rule at the review stage rather than only at the motion stage.

A Brady Review Workflow That Catches What the Line ADA Missed

The line ADA handing you the discovery dump is not the person who assembled it. In 51% of successful Brady claims, the withheld material never even reached the line prosecutor, and law enforcement accounts for 31% of all suppressions. Your review workflow has to assume the person producing discovery is a filter, not a fountain.

Start with the roster. Send an early letter or raise it at the first status hearing demanding identification of every member of the prosecution team: the arresting agency, the lab, the jail, dispatch, every detaining officer. McAward describes prosecutorial teams as "just a collection of parts that do not work well together," and you cannot spot an omission from a component you never named. In Los Angeles, the DA's office coordinates with nearly 100 separate law enforcement agencies; your case probably touches several.

Then tag every produced item as core Brady or impeachment material. This matters because the Jencks-trumps-Brady delay tactic depends on blurring the two. Three circuits (Fifth, Sixth, Ninth) let prosecutors hold impeachment material until after direct examination; three (Second, Third, D.C.) hold Brady requires earlier production. If you tag at intake, you can object with the right framing the day a late production lands, instead of untangling categories mid-trial.

Track the production method for each item, every time:

• Date received and in what form: full production, "viewer" access, or onsite-only review.
• Items reviewed under restricted access, logged in a spreadsheet so you can argue "effective use" was denied.
• Anything promised in open court but never delivered, with the hearing date.
• Each item tagged core-versus-impeachment, with the disclosure timing your circuit requires.

At the review stage, apply Kyles v. Whitley the way the Court wrote it: all undisclosed information gets considered collectively, not item by item. A single small withholding looks immaterial in isolation. Three small withholdings, taken together, can put the whole case in a different light, which is the actual standard. Your review has to cross-reference, because the government never will.

That collective-consideration discipline is where tooling earns its keep. Evidence Copilot runs auto-signals on every ingest, comparing each new file against evidence already in the case, so a report-versus-bodycam contradiction surfaces before anyone asks. Its Brady list research workflow then works in-case: each flagged withholding carries a citation to the exact transcript second or PDF page, and named-party records surface mentions of officers and witnesses across notes and signals. That is how a brady violation argument gets built: not from one dramatic document, but from a record of small, dated, verifiable gaps the prosecution team left behind.

Build the Record With Grounded Citations

A Brady motion lives or dies on materiality, and under Kyles v. Whitley materiality means a "reasonable probability" that the withheld evidence would have changed the outcome. That is not a sufficiency-of-the-evidence test. The defendant only has to show the suppressed information "can reasonably be taken to put the whole case in a different light." But you cannot put the whole case in a different light with vague references to "impeachment materials." You do it with a record.

So treat every entry on your Brady log as a citation job. When you identify a candidate Brady item, pin it to the exact source moment: the second on the bodycam where the officer's account diverges from the report, the page of the 911 call transcript where the caller's description contradicts the witness statement, the timestamp in the jail call where the timing falls apart. If the item lives only in your head as "the report is wrong about the sequence of events," it will not survive a materiality fight. If it lives as "report page 4, ¶ 3 vs. bodycam 00:41:17," it can.

The stakes of getting this right are not theoretical. In the largest Brady database ever assembled, covering 386 adjudicated violations from 2004 to 2022, 45% of cases involved suppressed evidence that could have impeached a critical prosecution witness. Impeachment material is exactly the kind of evidence courts weigh item by item, which is why Kyles matters so much: the Supreme Court requires that "all information not disclosed must be considered collectively, not item by item." Your record has to make the cumulative argument easy for the judge to make.

That means your Brady log should capture, for every item:

• The exact source citation: transcript timestamp, document page, or video frame.
• The category: core exculpatory versus impeachment material, since that distinction drives disclosure timing and the live circuit split on Jencks-versus-Brady timing.
• Who on the prosecution team generated it, because law enforcement is responsible for 31% of Brady violations, and in 51% of successful claims the material never even reached the line prosecutor.
• How it connects to the other withheld items, so the collective-consideration argument is built in from the start rather than reverse-engineered at the hearing.

This is also where grounded tooling earns its keep. In Evidence Copilot, every answer the AI evidence agent gives carries a clickable citation into the exact second of video or page of the PDF, and every output is marked AI-generated and pending human review. The auto-signals that compare new evidence against earlier files in the case surface report-versus-bodycam and statement-versus-call inconsistencies as candidate items for your log, each already tied to its source moment. You stay the attorney; the tool just keeps the citations honest.

If you want to see how this fits a Brady list research workflow end to end, Brady list research workflow walks through the full process. You can book a 20-minute walkthrough at evidencecopilot.com/book, or start on the Solo plan at $39/user/month with 15 media-processing hours included. Both are low-pressure options; pick whichever fits where your practice is right now.