AI-native CLM replaces manual setup

This is really a story about CLM moving from software that asked legal teams to prebuild the maze, into software that can navigate the maze itself. Older CLM systems worked best after teams manually mapped approval paths, tagged fields, and maintained repositories, which made them sticky once installed but slow to stand up and often narrow in day to day use. Spellbook is attacking that setup burden by letting AI read inbound contracts, route work, apply playbooks, and monitor signed agreements with less manual scaffolding.1 2 3 4

• Ironclad shows why the old model still got traction. It became durable once configured, first through services and later through a self serve workflow builder, reaching about $150M ARR in January 2025. The problem was not that workflow CLM had no value, it was that value came after a long implementation hill.2 4
• In practice, many teams used CLM more like a contract filing cabinet than an always on operating system. In procurement workflows, CLMs such as Ironclad and Icertis often sat as document centric repositories, while work still spilled across email, spreadsheets, compliance reviews, and vendor systems. That fragmentation is exactly where AI intake and routing matter most.3
• Spellbook is widening the job from single document review to matter level work. Associate handles diligence sets, financing packs, and cross document comparisons, while ACM pulls contracts from email, Teams, Slack, or Salesforce into review, negotiation, storage, and post signature monitoring. That shifts the product from lawyer copilot toward contract system of action.1 4

The next phase of the market is likely a rebuild of CLM around AI native intake, triage, and background monitoring, not just AI features bolted onto repositories. If Spellbook keeps reducing setup work while expanding from Word into upstream requests and downstream storage, it can compete for the larger system of record budget that legacy CLM vendors spent a decade establishing.1 4