Trial Preparation: Exhibits, Cross-Exam, and Jury Themes

Trial preparation in 2026 is where AI delivers some of its most dramatic time savings. A 6-week trial prep that used to consume the productive capacity of a partner, two associates, and a paralegal can now be done with the same team in 3 to 4 weeks at higher quality. The trick is knowing which trial-prep tasks AI handles well and which still require human craft.

This lesson walks through the trial prep tasks where AI is now indispensable, and the tasks where it is still mostly unhelpful.

What You'll Learn

• How to use AI to organize the trial binder and exhibit list
• AI-assisted cross-examination outlines from the record
• Generating and stress-testing jury themes
• Building demonstratives and opening/closing scaffolding

The AI-Leveraged Trial Binder

A modern trial binder is increasingly digital, indexed, and AI-searchable. Even if you bring printed binders to court, the underlying organization is built on a searchable platform.

A workable trial binder organization for 2026:

• Pleadings and orders — final operative pleadings, key orders, jury instructions
• Exhibits — pre-marked and indexed, AI-searchable
• Witness binders — one per witness with key documents and prior testimony
• Theme binders — one per case theme with the strongest evidence for each
• Cross-examination outlines — one per opposing witness
• Demonstratives — final approved demonstratives with sources

The platform you use is less important than the discipline of having a single, AI-searchable index of every document in the case. Relativity's aiR for Case Strategy and Everlaw's case prep features both excel here, but a well-organized SharePoint with embedded search can also work.

Exhibit Organization

AI excels at exhibit work because it is fundamentally a document-organization task.

A typical AI-assisted workflow:

1. Candidate exhibit generation. Ask the platform: "Based on the case theory and witness list, identify the top 75 documents from the production set most likely to be used as trial exhibits. For each, identify the sponsoring witness and the legal point it supports."

2. Foundation analysis. For each candidate, ask: "Identify the witness who can lay foundation for this document, and any authentication issues the other side might raise."

3. Objection prep. For each exhibit, ask: "List the three most likely objections (hearsay, relevance, prejudice) and the strongest responses."

4. Marking and indexing. AI generates the exhibit list in the format your court requires, with descriptions consistent with your local rules.

A senior associate or partner then reviews the entire exhibit list, but the heavy lifting of identification and foundation analysis is done in hours rather than days.

Cross-Examination Outlines

The deposition-prep workflow from earlier in this course is the foundation, but trial cross-examination has unique features.

A trial cross-examination outline focuses on:

• Locking in admissions that support your case theme
• Setting up impeachment with prior inconsistent statements
• Limiting damaging testimony through leading questions and exhibit control
• Avoiding rabbit holes — every question should serve the theme

AI helps with each of these. A useful prompt structure:

For the witness {witness_name}, prepare a trial cross-examination
outline organized by case theme. Use the witness's deposition
transcript, all produced documents authored by or sent to the
witness, and the operative complaint.

For each theme:
1. List the admissions you want to lock in (as questions in the form
   "isn't it true that...")
2. List the documents to mark and the order to mark them in
3. List any prior inconsistent statements available for impeachment,
   with deposition page and line cites
4. List topics to avoid

Use short, single-fact, leading questions.

Replace \{witness_name\} with the specific witness. The output is a usable first draft; the trial attorney edits it into the final outline.

Generating and Stress-Testing Jury Themes

Themes are the soul of a trial. AI cannot invent your theme — that is the trial lawyer's craft. But AI can stress-test the themes you have and help you choose between them.

A workable stress-test prompt:

Below are three candidate jury themes for our case. For each theme:
1. Identify the strongest evidence supporting it
2. Identify the strongest evidence against it
3. Identify the most damaging cross-examination opposing counsel
   could mount against it
4. Identify the jury demographics most likely to be receptive
5. Estimate the theme's resilience if the jury hears unexpected
   adverse evidence

The output is rarely a clear winner, but it surfaces hidden weaknesses before you commit. The decision remains yours.

Drafting Opening and Closing Scaffolds

Openings and closings should sound like you, not like an AI. But AI can produce a useful scaffold.

For opening, ask:

Draft a 10-minute opening statement scaffold organized by case
theme. Use only facts in the record. Use plain language. Avoid
legal jargon. Indicate where exhibits should be referenced.
Do not include any argument the court has excluded.

For closing, ask:

Draft a 15-minute closing argument scaffold tying our case
themes to the jury instructions provided. For each instruction
element, identify the evidence in the record that supports
our position.

The scaffold is a starting point. The trial lawyer rewrites it in their own voice. The point is to save the 6 to 8 hours of staring at a blank page.

Demonstratives

Trial graphics in 2026 are increasingly AI-assisted. A workflow:

1. Identify the 5 to 10 concepts the jury must understand.
2. For each, ask the platform: "Design a demonstrative exhibit that conveys this concept to a non-expert juror. Describe the visual format, the data points, and the source of the data."
3. Hand the descriptions to a trial graphics vendor or an in-house designer.
4. Have the AI critique the resulting drafts: "Identify any visual elements that could be misleading, any data not in evidence, and any improvements for clarity."

Most trial graphics teams in 2026 use AI for the first-draft concept generation and rely on humans for the actual design.

What AI Still Does Poorly

It is worth flagging the tasks AI is still bad at:

• Reading the jury in real time. No AI tool can sit in your peripheral vision during voir dire and tell you which juror is reacting.
• Witness handling. Pacing, tone, timing of impeachment — these are craft.
• Court relationships. Judicial preferences and courthouse practice norms still require humans who have practiced there.
• Strategic concessions. Deciding what to give up to win is a senior lawyer's call.

If you delegate any of these to AI, you will lose trials. Use AI to free up the human bandwidth for these tasks, not to replace them.

Cost and Time Estimates

A representative comparison for a two-week jury trial with one expert witness and 30 exhibits:

Task	2022 hours	2026 hours
Exhibit identification and foundation	40	8
Cross-examination outlines (3 witnesses)	60	18
Demonstratives concept work	25	8
Opening/closing first drafts	20	6
Trial binder organization	30	6
Total prep tasks	175	46

The savings flow to higher-value work: deeper witness prep, more time strategizing with the client, more practice on cross-examination.

Key Takeaways

• AI compresses trial prep dramatically without sacrificing quality, but only on the document-and-organization tasks.
• Use AI for exhibits, cross-examination outlines, jury theme stress-tests, and opening/closing scaffolds.
• Themes and craft come from the trial lawyer; AI tests them, it does not invent them.
• Save the time saved for the work AI cannot do — witness handling, courtroom presence, strategic judgment.
• Build a single, searchable, AI-aware trial binder from day one of trial prep; the discipline pays off in the courtroom.