Defining the New AI Legal Tech Wave
The new AI legal tech wave is the rapid shift from generic language models and off‑the‑shelf tools toward specialised, workflow‑aware, and often firm‑specific systems that embed large language models directly into legal processes spanning drafting, review, and knowledge management. This change is being driven by foundation model providers entering the legal vertical, elite law firms building their own platforms, and clients demanding secure, high‑impact automation for document‑heavy matters. Together, these forces are reshaping how legal work is organised, where value sits in the technology stack, and who captures it: big tech platforms, legal tech vendors, or law firms that build in‑house solutions. The emerging landscape suggests a contest over control of legal workflows rather than just access to models, with long‑standing vendors under pressure to defend their role between AI giants and their largest customers.
OpenAI’s Legal Vertical: From Models to Workflows
OpenAI’s creation of a dedicated legal vertical, led by Ironclad co‑founder Jason Boehmig, marks a clear move from being a general model provider to a focused player in AI legal tech disruption. Boehmig’s background is significant: Ironclad was one of the earliest legal tech vendors to build AI‑powered contract review and redlining on top of OpenAI models, shaping expectations for what generative AI could do in contracts. The move follows OpenAI’s own shift in message that “the model alone is no longer the product”, with attention turning to agents, workflow automation and industry‑specific applications. Legal is a natural target because it combines high‑value knowledge work with document‑heavy, repeatable tasks. OpenAI now joins Anthropic’s Claude for Legal and Microsoft’s emerging Legal Agent efforts, creating a trio of AI giants aiming at both in‑house counsel and law firms with increasingly tailored offerings.
Kirkland & Ellis’s $500m Custom AI Platform
Kirkland & Ellis has announced that it will spend USD 500m (approx. RM2,300m) over the next three to four years to build its own custom legal AI tools and services, funded directly from its revenue of USD 10.6bn (approx. RM48,760m). The firm’s goal is a broad platform that lawyers can use across their work, rather than a patchwork of point solutions. Undisclosed outside companies are helping build the technology, but they will not be allowed to sell the resulting tools to other firms, signalling that Kirkland wants firm‑specific differentiation. This stands in contrast to Freshfields’ deal with Anthropic, which allows the AI company to sell similar products more widely. Kirkland’s strategy fits a wider pattern: firms such as Simmons & Simmons and Allen & Gledhill are building proprietary platforms like Percy and A&GEL to keep data, inference, and governance firmly under their own control.
Custom Legal AI Tools and the Squeeze on Vendors
The combination of an OpenAI legal vertical and massive law firm AI investment changes the balance of power for legal tech vendors. As OpenAI, Anthropic and Microsoft expand legal‑specific offerings, they increasingly compete directly with contract lifecycle management (CLM) and productivity‑focused tools. Artificial Lawyer reports that “CLM and other contract‑related companies face a precipice” as AI giants push deeper into in‑house legal departments, which have fewer entrenched vendor relationships. At the same time, elite firms are building their own custom legal AI tools with outside engineering support that cannot be resold, taking potential product-market space off the table. Vendors still have defensible positions around “data fortresses” and systems that do more than productivity, such as document management, but generic AI assistants alone are no longer enough. To stay relevant, vendors will need sharper domain depth, unique data, or tight integration with the giants’ platforms.
What This Means for Law Firm Operations
For law firms, the rise of firm‑specific AI platforms signals a shift from experimenting with standalone assistants to redesigning core workflows around AI. Simmons & Simmons’ Percy shows what full adoption can look like: a generative AI platform built in‑house with its own legal inference engine that sits entirely inside the firm’s network and has reached 87% adoption among fee earners within a year. Allen & Gledhill’s on‑premise A&GEL platform was built specifically to satisfy strict client confidentiality expectations, showing how security and compliance are becoming design features, not afterthoughts. As more firms follow Kirkland & Ellis’s path, operations teams will need to blend model selection, data governance, and change management into a single strategy. Most will avoid locking into a single LLM provider, preferring a mix of OpenAI legal vertical offerings, Anthropic partners, and internal tools tuned for their own risk profile and client promises.