Reviewing contracts

How Clment AI reviews work — the four-step setup, the scopes available, the review strategy that sets the AI's posture, the staged results screen, and the decision model you use on every finding.

Updated 11 Aug 2026

A Clment review takes one or more contracts, runs them through a rulebook (sometimes known as playbooks), and produces a list of findings — one per clause the rulebook tells it to look at. This article is the launcher: it explains the four-step setup, points at the deeper articles on each step, and walks the staged results screen where you assess the findings and produce your outputs.

The four-step setup

From any contract page, click the review button top-right — Start AI Review / Compare on a contract that hasn’t been reviewed yet, Review / Compare once it has — to open the AI Review dialog. It steps through four things:

Documents — the primary version, any attached documents (each marked Review or Supporting), and ad-hoc context materials.
Coverage — whole document, or just what’s changed (tracked changes / a comparison with another version).
Rulebook — which rulebook to evaluate against, plus optional free-text instructions for this specific run.
Approach — the review strategy (how assertive the review is and how much lands in your redline), an optional executive summary, and the review depth.

Each step is worth understanding in its own right.

Step 1 — Documents

The primary version of the contract is the default target. If the contract has attachments, you decide each one’s mode:

Review — findings are raised against this document.
Supporting — the AI sees it for context but won’t raise findings against it.

You can also add Additional context — ad-hoc reference material for this review only (emails, policy snippets, precedent clauses) that isn’t itself reviewed.

This is how a single review can span multiple documents and pull in reference material. See Multi-document reviews for the full pattern.

Reviewing a scanned document? If the primary version is a low-quality scan, the setup shows a “Repair document (recommended)” prompt. Accepting it has Clment’s AI rebuild a clean, accurate copy of the text before the review runs, so findings come from what the contract actually says rather than a garbled read — see Uploading contracts for how scanned documents are handled.

Step 2 — Coverage

The coverage picker decides what the AI looks at within the chosen documents. It starts as two top-level cards:

Whole document — full rulebook review across the entire contract. The first-look choice.
Tracked changes or Comparison — review just what’s changed. Choosing this reveals two sub-cards:
- Tracked changes — analyse the <ins>/<del> markup already in a DOCX, with author and date filters.
- Compare with another version — diff this contract against another version or contract, and review the differences.

The two differential options (Tracked changes and Compare) unlock a polished side-by-side comparison view that links findings directly to the changes they’re about. See Review scopes — whole document, tracked changes, version comparison for which to pick when, and Comparing versions side-by-side for the comparison surface.

Step 3 — Rulebook

Pick which rulebook the AI should evaluate against. Optionally add review instructions — free-text guidance for this specific run that supplements the rulebook’s standing instructions. Useful for “focus on liability and indemnity, this isn’t a full pass” or “this is a renewal, flag anything that changed from the prior year”. If you don’t pick one of your own and the contract is a recognised type, the review uses the Clment Rulebook for that type — a proper, type-tailored review with no setup — and any instructions you add still layer on top.

Step 4 — Approach

Set the review strategy — the single most important choice in the setup. It answers “where are you with this contract?” and tunes how assertive the review is and how much lands in your redline by default:

Understand & Negotiate — understand the whole contract, with recommendations for negotiation. Ideal for a first pass. The default.
High-priority only — close to signing; just the things that could genuinely cause problems.
Strict — flag every deviation from the rulebook, however small.

The strategy never changes what is checked (that’s the rulebook) — only the posture toward what’s found. This step also has an Add an executive summary toggle (a plain-English overview for non-lawyers, alongside the detailed findings) and the review depth (Standard or Deep). See Choosing a review strategy for the full breakdown.

Running the review

Click Run review (or Run comparison for a version compare). For differential coverage the live change count appears on the scope card. Typical timings:

Whole document, single contract — 60–90 seconds.
Whole document, multi-document review — 2–4 minutes.
Tracked changes — usually faster than the whole-document equivalent because there’s less text to evaluate.
Compare with another version — proportional to the size of the diff, not the size of the documents.

Findings appear as they’re generated; you can start working through the early ones before the later ones finish.

The results screen — a staged journey

When the run finishes you land on Overview, the first of three tabs that pace the work out. A thorough review can produce 30+ findings; the screen is built so you never face them as one wall of detail.

Overview — the outcome at a glance, and the suggested path through it. Big tiles summarise the run: the risk level, the finding count, and how many checks Passed, raised Warnings, Failed, or were Out of scope (rulebook items that don’t apply to this document). Below them, four stages you can work in order:
1. Read the overall summary — the executive summary, commercial assessment, and negotiation priorities in one place.
2. Assess the findings — jumps to the Findings tab and tracks your progress (“12 of 32 assessed”).
3. Produce Outputs: Updated Contract · Share · Refine Rulebook — the report and sharing are available anytime; the updated contract unlocks once your selections have direction.
4. Close the review — record the outcome when you’re done.
The review’s configuration (documents, coverage, rulebook, strategy) and the strategy-alignment notes sit in collapsed panels at the bottom — there when you need them, out of the way when you don’t.
Findings — where you do the human pass, one card at a time (next section).
Finish up — the outputs: create the updated contract, export or share the report, apply rulebook refinements, and set assignment and status.

Clment remembers where you were — reopening a review you’ve started assessing takes you straight back to the findings you haven’t touched yet.

Assessing a finding

Each finding card walks you through the same four things:

The AI’s analysis — what it found, why it matters, and its recommendation, with the clause quoted. The card also shows the check result the Overview tiles count (Passed / Warning / Failed / Out of scope) — hover the icon for what it means.
Your call — Agree, Partially, or Disagree with the AI’s recommendation. Note that “Agree” means “agree with the recommendation” — not “agree with the clause as drafted”. If the AI recommends rewording a limitation-of-liability cap and you say Agree, you’re saying yes, reword it.
A note — optional free text for your reasoning; the card prompts “What would you change?” when you pick Partially or Disagree. Notes travel to the exported report, the activity log, and rulebook refinement.
What changes in your updated contract — one control settling both whether the clause changes and what it says: Use this fix (the AI’s drafted wording), Custom instructions (yours instead), or No change.

A finding set to change the document with nothing to go on — no AI-drafted fix and no custom instructions — shows as Needs direction, and the output step will point you back to it before generating.

The Findings tab keeps long lists manageable: filter chips (Not assessed, Needs direction), severity and change-status filters, a progress bar with a Next unassessed jump, and a grid view for scanning everything you’ve decided in one sortable table. See Understanding finding verdicts for the deeper version of the decision model.

Producing your outputs

The Finish up tab collects the outputs in one place:

Create updated contract — a Word document with tracked changes for every finding you set to change. For multi-document reviews, you get one per Review-mode document. See Generating redlines for the full story.
Report & sharing — copy or export the review report, or copy a deep-link for a colleague. Available from the moment the review finishes; you don’t have to assess everything first.
Rulebook refinements — apply the improvements your calls and notes suggested. See Refining a rulebook from review data.
Status & assignment — assign the review to a colleague, re-run it, or close it.

Closing the review

When you’re done, mark the review Closed — from the Status & assignment card on Finish up, or from the last stage on Overview. Closing is a one-off, review-level state: the review records who closed it, and it can be reopened at any time if there’s more to do. If the contract moves again, run a fresh review — each run gets its own REV-N identifier, so the closed review stays intact as the record of what was decided.

Reviews can be assigned to a colleague before closing if you want a second opinion — they’ll get an in-app notification and an email link with ?review=<id> so they land directly on the review.

Still have questions?

Instant article search