What this automation does
When a new contract arrives for review, this automation reads the full document and generates a structured summary: parties involved, effective date, term length, payment terms, liability caps, indemnification clauses, termination conditions, non-compete restrictions, and any unusual or non-standard clauses that need attention.
Legal review of contracts is expensive and slow — lawyers charge $200-500/hour and review can take days. This automation provides an instant first-pass analysis that catches 90% of the issues, letting your legal team focus on the flagged items rather than reading every word of every contract. It does not replace legal counsel but dramatically reduces the time they need to spend.
Tools you need
- Google Drive or Dropbox: Where contracts are uploaded for review
- OpenAI API (GPT-4) or Claude: Reads and analyzes contract text ($0.15-0.40 per contract)
- Make or Zapier: Triggers the review when a new contract is uploaded and delivers the analysis
How to set it up
Step 1: Create a 'Contracts for Review' folder in Google Drive. When a team member uploads a contract, the automation triggers. Extract the full text from the PDF — for simple PDFs, use text extraction; for scanned contracts, use OCR.
Step 2: Send the contract text to GPT-4 or Claude with a detailed prompt. Ask it to extract: parties and their roles, effective date and term, payment terms and amounts, liability limitations and caps, indemnification obligations, termination clauses and notice periods, non-compete and non-solicitation terms, confidentiality requirements, and governing law and jurisdiction.
Step 3: Ask the AI to also flag any clauses that are unusual, one-sided, or potentially risky. For each flagged clause, request a plain-English explanation of the risk and a suggested modification. Rate overall contract risk as low, medium, or high based on the number and severity of flagged items.
Step 4: Generate a one-page contract summary document in Google Docs. Include the extracted terms in a table, flagged clauses with risk explanations, and an overall risk assessment. Share the summary with the relevant team member and, if risk is medium or high, automatically notify your legal team via email or Slack.
Cost breakdown
| Item | Cost | Notes |
|---|---|---|
| OpenAI API or Claude | $15-$35/mo | ~$0.25 per contract at 60-140 contracts/mo |
| Make or Zapier | $15-$20/mo | Based on contract volume |
| PDF extraction/OCR | $0-$5/mo | For scanned or image-based contracts |
| Setup time | 35-60 min | One-time |
| Total monthly | $30-$60/mo | Saves 8+ hours/week of initial contract review |
Frequently asked questions
No, and it should not. AI provides an excellent first-pass analysis that surfaces the important terms and flags potential risks. But contract law is nuanced and context-dependent — a clause that is standard in one industry may be problematic in another. Use AI to save your lawyer time, not to replace their judgment. Think of it as a paralegal that reads everything instantly.
OpenAI and Anthropic's API data is not used for training and is encrypted in transit and at rest. For highly sensitive contracts, you can use Azure OpenAI or AWS Bedrock which offer additional data residency and compliance guarantees. Some organizations also run local models for maximum confidentiality.