The Digital Evidence Gap

The doctrine hasn't changed. The evidence has. Brady v. Maryland still requires disclosure of favorable, material evidence, and Giglio still extends that duty to impeachment material. What's new in 2026 is the sheer volume of digital evidence flowing into discovery, and the way favorable material can sit unnoticed inside it.

Aaron A. Pelley, a partner at Gleam Law, names the problem directly: the explosion of digital evidence lets favorable material hide in plain sight. A generic open-file policy can look like compliance on paper while the defense never actually finds the report-versus-bodycam contradiction that wins the suppression hearing.

The numbers back that up. Of 181 NYPD officers the CCRB found had made false official statements, the department disclosed the credibility finding for just 5. That's a disclosure rate near 2.8%. Having a Brady list and disclosing from it are two different jobs.

Why the gap? Under Kyles v. Whitley, the prosecution team includes police and investigators, and DOJ policy tells federal prosecutors to seek exculpatory and impeachment information from all of them. Decentralized evidence systems make that duty hard to discharge, and most violations trace to case-file problems rather than deliberate concealment.

For the defense, that reframes the workflow. The Brady battlefield is the discovery dump, not the police report. You need to systematically review:

• Bodycam and dashcam footage, checked line by line against the corresponding report
• Interrogation recordings and Miranda timing
• 911 audio and jail calls for statements that contradict the official narrative
• Officer disciplinary records, CCRB findings, and Brady-list status
• Informant and cooperating-witness deals documented anywhere in the file

That's the operational answer to "hidden in plain sight." A structured Brady list research workflow cross-references officer names against disciplinary records and prior inconsistent statements, so the impeachment material surfaces before the hearing instead of after the conviction.

Strickler Elements in a Digital Context

Proving a Brady violation requires you to satisfy a specific three-part test established in Strickler v. Greene. The doctrine has remained stable for decades, but the volume of digital evidence has shifted the operational burden entirely. You must demonstrate that evidence favorable to the defense was suppressed by the State and that this suppression prejudiced the outcome.

The first element focuses on the nature of the evidence. It must be favorable to the defense, which includes both exculpatory material and impeachment evidence. Under Giglio v. United States, this extends to any information casting substantial doubt on the accuracy of prosecution witnesses. The Supreme Court clarified that the jury was entitled to know about witness deals because credibility is central to the case. If the State fails to disclose such material, it violates the duty to present all material evidence to the jury.

The second element requires you to show the State suppressed the evidence. This is where the definition of the "prosecution team" becomes critical. Under Kyles v. Whitley, the prosecutor is responsible for favorable evidence held by police and investigators, even if the prosecutor’s office never personally possessed it. The DOJ Justice Manual explicitly includes federal, state, and local law enforcement in this obligation. This means a violation occurs when a detective withholds a bodycam file or an internal affairs report, regardless of whether the prosecutor ever reviewed it.

The third element is prejudice. You must show a reasonable probability that the result would have been different had the evidence been disclosed. This is a high bar, but it is the necessary conclusion of the first two elements. Crucially, you do not need to prove bad faith. The duty to disclose applies even if the failure was a matter of negligence rather than intent. Most violations arise from decentralized evidence systems where police hold materials separately from prosecutors, creating "case file problems" rather than intentional concealment.

When building your motion, focus on these structural points:

• Identify the specific favorable evidence, whether exculpatory or impeaching.
• Establish that the evidence was within the control of the prosecution team, including law enforcement.
• Demonstrate materiality by showing how the omission affected the trial’s outcome.
• Document the timeline of suppression to show when the State should have disclosed.

Negligence is sufficient for liability, so avoid wasting pages arguing for bad faith. Instead, frame the motion around suppression and materiality. A clean, evidence-based argument that walks the court through the Strickler elements is more persuasive than rhetoric-heavy misconduct claims. In the digital context, "hidden in plain sight" evidence often exists in unreviewed discovery dumps. A structured Brady list research workflow helps cross-reference officer names against disciplinary records and prior inconsistent statements. This operational precision ensures you do not miss favorable material buried in large file sets.

The gap between policy and practice is stark. Data from the New York CCRB shows that of 181 officers found to have made false official statements, the NYPD disclosed the credibility finding for only five. This illustrates the risk of relying on generic open-file policies. Specific, targeted requests convert paper compliance into actual disclosure. You must treat the discovery dump as the Brady battlefield, not just the police report. Systematic review of bodycam, dashcam, and interrogation footage is essential to finding inconsistencies that prove suppression.

Evidence Copilot supports this process by providing grounded evidence review for criminal-defense teams. Every AI answer cites the exact source moment it came from, allowing you to verify claims against the original video or document. This transparency ensures your motion is built on verifiable facts, not assumptions. The platform’s focus on human review and immutable originals aligns with the rigor required for a successful Brady motion. You need tools that help you find the truth, not just process the files.

Targeted Disclosure Strategies

A generic "any and all exculpatory evidence" request gets you a discovery dump, not disclosure. The data backs this up: when the NYPD's CCRB found 181 officers had made false official statements, the department disclosed those credibility findings for only 5 of them, according to NACOLE's Brady and Giglio analysis. Generic requests don't surface that material. Specific ones do.

The reason is structural. Most Brady failures come from decentralized evidence systems where police hold material separately from prosecutors, what practitioners call "case file problems" rather than deliberate concealment. Under Kyles v. Whitley, the prosecution team includes police and investigators, so the State answers for what the arresting officer never handed over. But a vague demand gives a decentralized system no map. Targeted demands give it one.

Name the categories. Name the officers. When you invoke Giglio explicitly and identify the testifying officer by name, you convert paper compliance into an obligation the prosecutor can't wave through. Your demand list should cover:

• Informant and cooperating-witness deals, benefits, and expectations of leniency, the exact category that produced the Giglio violation itself
• Prior inconsistent statements, sworn and unsworn, from every witness the State intends to call
• Officer Brady-list status and any disciplinary or internal-affairs records involving false-statement allegations
• CCRB or civilian-oversight findings touching your testifying officers
• Lab and forensic results, including any contrary or excluded findings

The CCRB numbers explain why the officer-specific framing matters. A prosecutor's office can honestly say "we asked the department" and still miss a credibility finding, because the list exists but the disclosure protocol behind it doesn't work. NACOLE notes that some Brady lists and protocols "have been criticized as inadequate." Your demand has to assume the list won't reach the courtroom on its own.

Then work the dump itself. Favorable material increasingly hides in plain sight inside bodycam files, 911 audio, and jail calls that arrive as unsearchable formats with buried timestamps. Aaron Pelley's observation is the operational diagnosis: if the defense can't reasonably find or understand the favorable evidence, the State may have "complied on paper while still undermining Brady in practice." Comparing each recording against the corresponding police report is where report-versus-footage inconsistencies surface, and a structured Brady list research workflow that cross-references officer names against disciplinary records and prior statements is how defense teams make that comparison systematic rather than aspirational. Tools like Evidence Copilot exist to make that cross-referencing citable, with every answer tied to the exact moment in the source file.

Finally, build the record as you go. Log every request, every production, and every gap, because the Strickler test requires you to prove suppression, and negligence alone is enough to establish it. A clean timeline of who knew what, and when, wins Brady motions.

Workflow for Digital Discovery Review

Most Brady material in 2026 doesn't arrive in a separate envelope. It arrives buried in the discovery dump, inside hours of bodycam footage, 911 audio, and jail calls that no one has transcribed. As Aaron Pelley put it in a recent discussion of Brady compliance, favorable evidence can now be "hidden in plain sight," and if the defense can't reasonably find it, the State may have complied on paper while undermining Brady in practice (Rev.com Brady guide).

The fix is a systematic cross-referencing workflow, not a heroic all-nighter before the preliminary hearing.

Start with the recordings, then the reports. For each police report, pull the corresponding bodycam video, the 911 call that triggered the stop, and any jail calls from the reporting officer's witnesses. Transcribe everything with word-level timestamps and speaker diarization so you can quote precisely. Then compare, line by line, what the report claims against what the recording captures.

What you're hunting for are the inconsistencies that become impeachment material:

• Report says the suspect consented to a search; the bodycam audio shows no consent question was ever asked.
• The report describes a witness "confirming" the officer's account; the 911 call transcript tells a different story.
• Timeline gaps: the report's sequence of events doesn't match the timestamps on the footage.
• Statements attributed to defendants or witnesses that don't appear anywhere in the recorded interactions.

Every inconsistency you find is potential Giglio material, since impeachment evidence casting doubt on the accuracy of prosecution evidence falls within the disclosure duty (DOJ Justice Manual §9-5.000). And remember: the duty applies even when the failure was negligent rather than intentional, so you don't need to prove bad faith, only suppression and materiality.

Manual comparison breaks down at scale. This is where a grounded review workspace earns its keep: upload the full dump, get timestamped transcripts and OCR'd report text, and let the tool surface candidate inconsistencies (report vs. bodycam, statement vs. call audio) as review signals before you've thought to ask. Evidence Copilot's auto-signals do exactly this on ingest, and every flagged quote links back to the exact second of video or page of the PDF, so you can verify the machine's work against the original.

Two guardrails keep this defensible. First, treat every AI-flagged inconsistency as a lead requiring human verification; the tool organizes the comparison, you make the credibility call. Second, log what you find and when. If a report-vs-bodycam contradiction never gets disclosed, your documented review becomes the record supporting a later Brady motion.

The NYPD data shows why this matters: of 181 officers found to have made false official statements, the department disclosed the credibility finding for only 5 (NACOLE). Nobody is going to hand you the inconsistency. You have to build it yourself, recording by recording, report by report.