1. The Scenario
Your client, a Fortune 500 technology company, is considering acquiring a Series B AI startup for approximately $85 million. The startup's primary value lies in its intellectual property: 12 patent applications, a proprietary machine learning model, and 47 software licenses. The deal team needs an IP due diligence summary within 48 hours.
You have received a virtual data room with 280+ documents: patent applications, employment agreements with IP assignment clauses, third-party license agreements, open source usage records, and prior art searches.
Key details:
- Target: Series B AI startup (~50 employees)
- IP portfolio: 12 patent applications, ML model, software licenses
- Data room: 280+ documents across 15 folders
- Timeline: 48 hours to executive summary
- Key risk area: Open source license compliance and IP ownership chain
- Deal value: ~$85M, with IP representing ~60% of the valuation basis
2. The Traditional Approach
IP due diligence on a tech startup is notoriously document-intensive. The traditional approach requires multiple attorneys across specialties:
Document review and indexing (8-12 hours)
Review every document in the data room. Create an inventory of all IP assets. Index patent applications, licenses, and employment agreements. Flag missing documents.
Patent analysis (4-6 hours)
Review each patent application for scope, status, and potential issues. Check assignment chains. Identify continuation or provisional filings. Note any prosecution deadlines.
License and open source review (4-6 hours)
Review every third-party license for change-of-control triggers, restrictions on assignment, and scope limitations. Audit open source usage for copyleft or viral license risks.
Ownership chain analysis (3-4 hours)
Verify that all IP is properly assigned from founders and employees to the company. Check for gaps in assignment chains, contractor IP issues, and pre-incorporation contributions.
Summary drafting and review (4-6 hours)
Synthesize findings into an executive summary with risk matrix. Internal review and quality check across the team.
Traditional Time Estimate
23-34 hours of total attorney time (typically spread across 2-3 attorneys), over 2-4 days.
3. The AI-Assisted Approach
AI dramatically accelerates the initial document review and extraction phase. The approach: use AI for document-level extraction and categorization, then apply expert judgment to the synthesized results.
Batch Document Processing (30-60 min)
Upload documents in logical batches to an AI with a large context window. Group by type: patent applications together, employment agreements together, licenses together. Run the extraction prompt on each batch.
Run the Due Diligence Prompt (10 min)
Use the comprehensive prompt below on the full document set (or on the extracted summaries from Step 1). It instructs the AI to produce a structured due diligence summary with risk categorization.
Targeted Deep-Dives on Flagged Issues (2-3 hours)
For each issue the AI flags as HIGH or CRITICAL risk, go back to the source document and review it yourself. Use the AI to cross-reference: "Does any employment agreement in the data room lack a patent assignment clause? List specific employees."
Gap Analysis and Missing Document Identification (30 min)
Ask the AI: "Based on the documents reviewed, what standard due diligence items are missing from this data room?" Cross-reference against your firm's DD checklist. Send a document request to the target.
Draft and Finalize Executive Summary (1-2 hours)
Use the AI's structured output as the framework for your executive summary. Add your professional judgment on deal-specific implications, quantify risks where possible, and ensure recommendations are actionable for the deal team.
4. The Prompt
This prompt is designed for use with a large-context AI (Claude with 200K context is ideal for document-heavy DD). Upload the documents alongside the prompt. For data rooms exceeding the context window, process in batches by document type.
Complete Prompt — IP Due Diligence Summary
You are a senior M&A attorney conducting intellectual property due diligence on a target company (an AI startup) for a potential acquisition. The acquirer is a Fortune 500 technology company. The deal is valued at approximately $85 million, with IP representing about 60% of the valuation basis.
Review the attached documents and produce a comprehensive IP due diligence summary with the following structure:
## 1. IP ASSET INVENTORY
Create a complete table of all identified IP assets:
| Asset Type | Description | Status | Filing/Registration Date | Jurisdiction | Assigned Owner | Notes |
Include: patents, patent applications, trademarks, copyrights, trade secrets (if documented), domain names, and any other IP referenced in the documents.
## 2. OWNERSHIP CHAIN ANALYSIS
For each material IP asset, trace the chain of title from creation to current ownership:
- Identify the original creator/inventor
- Verify assignment documentation (employment agreements, contractor agreements, founder assignments)
- Flag any gaps in the assignment chain
- Note any IP created before the company's incorporation
- Identify any employees or contractors whose agreements are missing from the data room
## 3. PATENT PORTFOLIO ANALYSIS
For each patent application:
- Application number, filing date, and current status
- Brief description of claimed invention
- Prosecution history summary (if available)
- Upcoming deadlines (office action responses, maintenance fees)
- Preliminary assessment of scope and potential value
## 4. LICENSE AND ENCUMBRANCE REVIEW
For each third-party license:
- Licensor, scope, and key restrictions
- Change of control provisions (will the license survive the acquisition?)
- Exclusivity provisions
- Termination triggers
- Financial obligations (royalties, minimum payments)
For open source components:
- License type (permissive vs. copyleft)
- Compliance status
- Any copyleft licenses (GPL, LGPL, AGPL) that could affect proprietary code
- Risk assessment for each copyleft component
## 5. RISK MATRIX
| Risk | Severity (Critical/High/Moderate/Low) | Impact on Deal | Recommended Action |
## 6. MISSING DOCUMENTS AND INFORMATION GAPS
List any documents or information that should be in the data room but are missing. For each, explain why it matters and what the absence could indicate.
## 7. EXECUTIVE SUMMARY AND RECOMMENDATIONS
- Overall assessment of the IP portfolio's strength and value
- Top 5 risks ranked by potential deal impact
- Recommended conditions or representations to include in the acquisition agreement
- Items requiring follow-up before closing
Be thorough, specific, and flag anything that could affect the deal valuation or create post-closing liability.Why This Prompt Works
Deal context
Specifying the deal value and IP valuation basis tells the AI the stakes involved, producing more careful and thorough analysis. The AI calibrates its risk assessments to the deal size.
Structured tables
Requesting tables for the IP inventory and risk matrix produces output that can be directly inserted into a due diligence report or shared with the deal team in a spreadsheet format.
Open source focus
For tech acquisitions, open source compliance is often the most complex and time-consuming part of IP DD. Explicitly asking for copyleft analysis ensures this critical area is covered.
5. The Review — Evaluating AI Output
Due diligence directly affects deal pricing and risk allocation. The AI's output must be rigorously verified before it informs any deal decision.
Verify the IP Inventory Against Source Documents
Confirm that every asset in the AI's inventory corresponds to a real document in the data room. Check for assets the AI missed. Verify patent application numbers against the USPTO database. An incorrect inventory could lead to overpaying for assets that do not exist or missing assets that should have been evaluated.
Spot-Check Assignment Chains
For the most critical IP assets (the patent applications and the core ML model), manually trace the assignment chain from creator to company. Verify that the AI correctly identified all inventors and that assignments are properly executed. Ownership gaps can be deal-killers.
Review Change-of-Control Provisions Yourself
For any license the AI flags as having a change-of-control provision, read the actual language. The nuances of these provisions (consent required vs. notification only, automatic termination vs. optional) have enormous implications for whether the license survives the acquisition.
Validate Open Source Analysis
Run the open source software list against a dedicated scanning tool (Black Duck, FOSSA, or similar) if available. AI may miss components or misidentify license types. Copyleft contamination of proprietary code is a serious risk that requires specialist review.
6. The Result — Time & Quality Comparison
Traditional Approach
23-34 hrs
Total attorney hours across a 2-3 person team over 2-4 days
AI-Assisted Approach
8-12 hrs
AI-assisted extraction and analysis with targeted human deep-dives
Where AI Adds the Most Value in DD
The biggest time savings come from the document extraction phase. Reading 280 documents, indexing their contents, and creating structured summaries is work that AI performs in minutes rather than days. This alone can reduce the initial review phase from 8-12 hours to under 1 hour.
However, the expert analysis phase cannot be shortened. Assignment chain verification, change-of-control assessment, and open source risk evaluation require legal expertise that AI cannot replace. The AI gets you to the analysis phase faster — it does not replace the analysis itself.
Try Another Application
Due diligence is just one M&A workflow where AI accelerates results. Explore contract review, legal research, and client communication.
Ready for structured learning? Explore the Learning Program →
Comments
Loading comments...