A legal team reviewing a standard vendor contract spends 3-4 hours reading through it carefully. They’re looking for problematic clauses, missing protections, unfavorable terms, and anything that deviates from their standard positions. It’s detailed, important work — and it’s the same set of checks every time.
When the team is reviewing 10-15 contracts a month, the math breaks. Either review quality drops (skim the boilerplate, focus on the unusual sections) or turnaround time extends (contracts sit in the queue for days). Neither outcome is acceptable when a missed liability clause can cost the company millions.
The Contract Review Pipeline
The Legal starter pack includes a workflow called Contract Review Pipeline. It reads a contract and produces a structured risk assessment with flagged clauses and recommendations.
-
Contract Summary — The AI reads the full contract and produces a structured summary: parties, effective date, term length, key obligations for each party, termination provisions, financial terms, and governing law. This gives the reviewer a quick orientation before diving into specifics.
-
Clause Review — The detailed analysis step. The AI evaluates each section against common risk criteria: indemnification scope, liability caps, IP ownership, data protection obligations, non-compete breadth, change-of-control provisions, and warranty disclaimers. For each flagged clause, the output includes: the clause text, the risk category, severity (high/medium/low), and a specific concern explaining why it’s flagged.
-
Condition: If High Risk — When any clause is flagged as high risk, the workflow routes the contract for senior review.
-
Legal Review (Approval Gate) — The workflow pauses and sends the full analysis — summary, flagged clauses, and risk assessment — to the assigned reviewer. The reviewer can approve the contract, reject it, or request specific redlines.
What the AI catches
The clause review recipe is configured with your organization’s standard positions. You define what “acceptable” looks like for each clause type:
- Indemnification: Your standard is mutual indemnification with a cap. The AI flags one-sided indemnification or uncapped liability.
- IP ownership: Your standard is that each party retains pre-existing IP. The AI flags broad IP assignment clauses.
- Data protection: Your standard requires specific data handling commitments. The AI flags contracts with no data protection provisions or that permit unrestricted data sharing.
- Termination: Your standard is mutual termination for convenience with 30 days notice. The AI flags contracts with no termination right or extended lock-in periods.
These standards live in the recipe’s prompt template. When your positions change — say, you adopt a new data protection standard after a regulatory update — you update the prompt once and every future review uses the new criteria.
What the lawyer still does
The AI performs the first pass. The lawyer makes the decisions.
- Risk tolerance is contextual. A one-sided indemnification clause in a $10K vendor contract is a different conversation than the same clause in a $2M partnership agreement. The AI flags both. The lawyer decides which one to push back on.
- Negotiation strategy is human. The AI identifies that the liability cap is too low. The lawyer decides whether to negotiate the cap, trade it for a concession elsewhere, or accept it because the vendor won’t budge and the deal is worth it.
- Precedent and relationships matter. The AI doesn’t know that this vendor accepted your standard terms last time, or that your CEO has a personal relationship with their CEO. The lawyer brings that context to the review.
The value isn’t replacing the lawyer’s judgment. It’s ensuring that every clause gets reviewed with the same thoroughness, regardless of how many contracts are in the queue.
Compliance and regulatory workflows
The Legal pack also includes a Compliance Audit workflow that checks your existing policies against current regulatory requirements. It’s particularly useful after regulatory changes — feed in the new regulation and your current privacy policy, and the AI identifies gaps and suggests specific updates.
The New Vendor Legal workflow combines due diligence checks with contract review: assess the vendor’s legal standing, review their standard contract, and generate a requirements checklist, all in one pass.
The full Legal pack
The Contract Review Pipeline is one of three workflows in the Legal starter pack:
- Compliance Audit — Policy review, gap analysis against regulations, and prioritized action plan
- New Vendor Legal — Due diligence, contract review, and requirements checklist for new vendor relationships
The standalone recipes handle NDA generation, privacy policy review, terms of service updates, regulatory change impact analysis, and board resolution drafting.