Legal AI commoditization and workflow ownership
Healthcare company associate GC on where legal AI products break down
The real risk for legal AI point solutions is that model quality is converging faster than workflow ownership. In-house buyers already view Harvey and Legora as expensive layers on top of enterprise ChatGPT plus Westlaw, while large firms increasingly buy them in small, targeted seat blocks rather than as firm-wide operating systems. That means the durable winners are more likely to control documents, research corpora, and daily workflow, not just provide a better chat interface.
-
For lean in-house teams, the missing piece is not another prompt box, it is software that can ingest a company’s own contracts, build a usable playbook, flag deviations in third-party paper, and do it inside the CLM flow. Without that, buyers say existing ChatGPT and Westlaw setups cover most of the value at lower cost.
-
Inside large firms, Harvey and Legora are proving stickier as practice group tools than as standalone platforms. Firms are negotiating 5 to 20 seat deployments, hot-swapping licenses, and buying based on client pull, geography, and specific workflow fit. That looks more like a premium application layer than a winner take all legal operating system.
-
The companies with the strongest structural position are the ones that already own the system of record or the underlying legal corpus. Clio is pairing practice management with vLex’s global research database, Ironclad already owns contract workflow, and Thomson Reuters has CoCounsel plus Westlaw. Those bundles are harder to displace than a standalone AI assistant.
From here, the market is likely to split in two. General legal copilots will keep compressing into lower priced features, while the highest value accrues to products that become the place where contracts enter, get reviewed, move through approval, and connect to trusted legal data. That is where standalone assistants either mature into real infrastructure or get absorbed by broader platforms.