Checkbox Launches 'First Pass' — AI Agent That Reviews Contracts Before Lawyers Do

By Hector Herrera | September 19, 2026

Checkbox launched First Pass on September 16, an AI agent that reviews contracts, invoices, and incoming legal requests, then routes, resolves, or flags them — before a trained attorney ever touches the work. More than 100 enterprise legal teams, including SAP, PepsiCo, and BILL, are already running the platform in production. The launch is a strategic bet that the highest-value AI intervention in a legal department isn't at the drafting stage but at the intake layer, where routine volume first enters the queue.

What First Pass Does Differently

Earlier legal AI tools assumed a lawyer was still reviewing every document — they made the lawyer faster. First Pass is designed to reduce the number of documents that require a lawyer at all.

The system is context-aware: it knows the organization's standard contract positions, risk thresholds, and routing logic. That means it can:

• Auto-approve low-risk vendor agreements that fall within pre-negotiated parameters
• Flag non-standard terms for attorney review with a summary of what's unusual and why
• Return routine requests to the requestor with self-service guidance, skipping the legal queue entirely

The change in unit economics matters. Instead of AI making each lawyer faster, AI reduces the count of legal actions requiring a lawyer. That shifts the value proposition from productivity to capacity — a different conversation with legal department heads and CFOs.

Why Enterprise Legal Teams Are Buying

In-house legal departments face a persistent volume problem: demand for contract review grows with the business, but headcount budgets don't keep pace. Most large legal teams handle thousands of contract requests annually; the majority are routine. First Pass targets that routine volume directly.

Checkbox already integrates with Salesforce, ServiceNow, and Microsoft 365 — a distribution approach that reduces friction for enterprise buyers evaluating the platform against existing CLM (contract lifecycle management) systems they already operate.

What to Watch

The commercial pressure point is whether AI intake tools become standalone purchases or get absorbed as features into dominant CLM platforms like Ironclad, Icertis, and Conga. Checkbox's existing integrations and named enterprise client list give it a credible competitive position, but if the major CLM vendors ship native intake AI, the standalone category will face consolidation pressure. The next 12 months will clarify whether intake AI is a product category or a feature.