Not based on a specific customer deployment. The company, figures, decisions, and outcomes below are a constructed example that shows how Intended’s mechanism applies to this kind of problem. They are not measured results from a named customer. Where we publish a real, attributed customer outcome, we will say so explicitly.
Governing AI-Driven Contract Review at a Law Firm
Consider a large law firm whose AI agents handle initial contract review, risk-clause identification, redline suggestions, and routing to specialty attorneys across multiple practice areas.
01 · the challenge
The kind of problem this addresses.
02 · how it works
See the difference.
Contract uploaded for review
Master Services Agreement, 42 pages
AI generates redline suggestions
14 recommended changes, 3 contain errors
Sent directly to client
No attorney review of AI recommendations
Client relies on flawed advice
Material liability clause missed
03 · the solution
What they deployed.
- — Policy rules for contract review, redlines, and client communications
- — Policy requires attorney review for all AI recommendations on contracts > $500K
- — The agent sends its own confidence score with each request; policy escalates anything below 85% to a senior partner
- — Escalations name the approver role for the practice area (M&A, IP, employment, commercial)
- — Every recommendation's decision and the approving attorney are recorded in a tamper-evident audit trail
- — The agent calls Intended through the SDK before it posts redlines to the document management system or sends anything to a client
04 · implementation
From zero to governed.
Week 1
Map
Catalogue the AI contract-review workflows across the M&A, IP, employment, and commercial practices.
Week 2
Configure
Define review thresholds by contract value, clause type, and practice area. Check them with policy simulation.
Week 3
Integrate
Add the SDK call ahead of the agents' document-management and client-communication actions. Give attorneys access to the approval queue.
Week 4
Enforce
Enable enforcement. High-value recommendations now wait for the matched attorney before they reach a client.
05 · illustrative outcomes
What this is designed to deliver.
Modeled figures for this scenario — what the workflow above is built to achieve, not measured results from a named customer.
0
Unreviewed AI advice to clients
High-value contracts require attorney sign-off by policy
0%
Recommendations with provenance
AI vs. attorney attribution on every change
0%
High-value reviews escalated
Routed to the matched specialty attorney
0
Policy outcomes
Allow, escalate to an attorney, or deny
06 · decision replay
Example decisions, full trace.
Sample decision records that show the shape of the evidence Intended produces. Illustrative, not drawn from a live customer’s logs.
legal.contract.review-recommendationRISK: 18/100ALLOWAI reviews NDA template for startup client, 6 pages, standard terms
Resolved by: Policy: standard NDA templates auto-approved with boilerplate check
legal.contract.review-recommendationRISK: 82/100ESCALATEAI reviews $8.2M acquisition agreement, identifies 14 risk clauses
Resolved by: M&A Partner (approved 11 of 14 recommendations)
legal.contract.redline-generationRISK: 56/100ESCALATEAI generates redline for employment agreement, 3 non-compete clauses flagged
Resolved by: Employment Law Associate (reviewed and approved)
legal.contract.client-communicationRISK: 64/100ESCALATEAI drafts client memo summarizing contract risk assessment
Resolved by: Supervising Partner (reviewed and approved the memo)
the takeaway
AI contract review is powerful, but sending unreviewed AI recommendations to clients is a malpractice risk. Authority gating routes every recommendation to the right level of attorney oversight based on contract value and clause type — and records who approved what — while routine reviews keep moving.
Why this pattern matters — not a customer quote.
Start governing AI legal operations
Free to start. See every AI decision from day one.