Artificial Intelligence · Pre-Seed Stage Due Diligence

Artificial Intelligence Pre-Seed Due Diligence Checklist

The complete Pre-Seed due diligence checklist for Artificial Intelligence startups. Prepare your data room, anticipate every investor request, and close your round faster.

All Checklists

1–2 weeks

Typical DD Timeline

DD Categories Covered

50+

Checklist Items Total

Artificial Intelligence Regulatory DD Context

AI regulation is rapidly evolving (EU AI Act, US EO on AI). Investors will assess regulatory exposure by use case. Document your intended use cases and risk tier classification explicitly.

Data Room Essentials for Pre-Seed

Documents to have ready before DD begins

Incorporation documents
Founder IP assignment agreements
Cap table (Carta or equity schedule)
Bank statements (3 months)
Any signed LOIs or customer contracts
Pitch deck and financial model

LEGALLegal Due Diligence

Basic corporate formation documents, founder IP assignment, and cap table verification. Investors rarely hire outside counsel at pre-seed — but they will verify these exist.

General Legal Items

Certificate of Incorporation and all amendments
Bylaws and any shareholder agreements
Cap table with fully diluted ownership by class
All convertible instruments (SAFEs, notes, warrants)
Option pool documentation and all grants
Founder IP assignment agreements
Co-founder agreements and vesting schedules
Material third-party agreements

Artificial Intelligence-Specific Legal Items

Training data licensing agreements and provenance documentation
IP ownership of model weights and outputs
Employee IP assignment agreements (especially for researchers)
AI output liability and indemnification clauses in customer contracts
Export control compliance for dual-use AI technologies

FINANCIALFinancial Due Diligence

Bank statements for 3 months, any existing revenue documentation, and a basic projection model. Financial audit is not expected.

General Financial Items

Financial statements (P&L, balance sheet, cash flow)
Bank statements (12+ months)
Accounts receivable aging report
Accounts payable aging report
Payroll records and headcount history
Revenue recognition policy documentation
Tax returns and any outstanding tax obligations
Insurance policies (D&O, E&O, general liability)

Artificial Intelligence-Specific Financial Items

GPU / compute cost per inference and margin trajectory
Revenue attribution by model version
Data labeling cost and quality assurance spend
Research and development capitalization policy
Customer contract structure (usage-based vs. seat-based)

PRODUCTProduct and Technical Due Diligence

Model card or fact sheet for each production model
Bias and fairness audit results
Hallucination rate and accuracy benchmarks vs. competitors
Data pipeline documentation and lineage
Model versioning and rollback procedures

MARKETMarket Due Diligence

Benchmark performance vs. frontier models (GPT-4, Claude, Gemini)
Customer use case concentration risk
Evaluation of competitive moat defensibility as models commoditize
Enterprise customer lock-in mechanisms

TEAMTeam Due Diligence

LinkedIn profile verification and informal reference calls with 2–3 professional contacts. Background check is increasingly common.

LinkedIn profiles for all founders and officers
Employment agreements for key personnel
Equity grant history and vesting schedules
Any prior employer IP invention disclosure forms
Non-compete and non-solicitation agreements
Professional reference contacts (3+ per founder)
Advisory board agreements and equity grants
Organizational chart as of current date

Frequently Asked Questions

How long does Pre-Seed due diligence typically take for Artificial Intelligence startups?

Artificial Intelligence Pre-Seed due diligence typically takes 1–2 weeks. Basic corporate formation documents, founder IP assignment, and cap table verification. Investors rarely hire outside counsel at pre-seed — but they will verify these exist. Having a complete data room ready before DD kicks off can reduce this timeline by 30–50%.

What Artificial Intelligence-specific items do investors review during Pre-Seed DD?

For Artificial Intelligence at the Pre-Seed stage, investors focus heavily on: Training data licensing agreements and provenance documentation, IP ownership of model weights and outputs, and GPU / compute cost per inference and margin trajectory, Revenue attribution by model version. AI regulation is rapidly evolving (EU AI Act, US EO on AI). Investors will assess regulatory exposure by use case. Document your intended use cases and risk tier classification explicitly.

What should I put in my data room?

Your Pre-Seed data room should include: Incorporation documents; Founder IP assignment agreements; Cap table (Carta or equity schedule); Bank statements (3 months); Any signed LOIs or customer contracts; Pitch deck and financial model. Use a structured folder system that mirrors investor expectations — most institutional investors use a standard folder taxonomy.

What are the most common due diligence deal-killers?

The five most common DD deal-killers are: (1) undisclosed founder litigation or criminal history, (2) IP ownership gaps — particularly for university-origin technology, (3) customer contract terms that prevent assignment on change of control, (4) cap table math errors or undocumented equity grants, and (5) financial restatements required after revenue recognition review.

What team due diligence should I expect at the Pre-Seed stage?

LinkedIn profile verification and informal reference calls with 2–3 professional contacts. Background check is increasingly common.

Download the Full Checklist

Get the Artificial Intelligence Pre-Seed due diligence checklist as a Google Sheets or Notion template. Track completion status for every item in your data room.

Includes data room folder template, investor question tracker, and reference FAQ guide

Other Artificial Intelligence Stages

Seed Series A Series B Growth