What the New Legal AI Power Shift Looks Like
The current shift in legal technology is a consolidation of power in which large law firms and major AI companies build or control legal AI tools themselves, reducing reliance on smaller vendors and pushing the market toward enterprise-grade, end‑to‑end platforms tightly integrated into core legal workflows. Kirkland & Ellis has become the emblem of this trend, announcing plans to spend USD 500m (approx. RM2,300m) from its own USD 10.6bn (approx. RM48,760m) in annual revenue over three to four years on custom AI tools. The goal is a broad platform that lawyers can use across matters instead of juggling separate applications, and outside developers involved are barred from selling the same technology to rival firms. At the same time, OpenAI and Anthropic are moving beyond general models into legal‑specific workflows, turning the OpenAI legal vertical into a direct competitor to traditional legal tech startups.
Big Law’s Build Strategy: From Experiments to Enterprise Platforms
Kirkland’s investment signals that law firm AI adoption at the top end is shifting from pilots with third‑party tools to strategic, firm‑wide platforms. By funding a proprietary system and blocking its external partners from reselling it, Kirkland is turning AI into a competitive asset, not a commodity subscription. Other firms are following similar paths. Simmons & Simmons created Percy, a generative AI platform built entirely in‑house using its own large language model team, reaching 87% adoption among fee earners within a year. Allen & Gledhill’s A&GEL runs as a custom platform hosted on‑premise to satisfy strict client confidentiality demands. These efforts show legal technology investment moving toward tightly controlled infrastructure, where data, workflows, and model configurations stay inside the firm, and outside vendors play a supporting role rather than owning the primary interface lawyers use every day.
OpenAI’s Legal Vertical: From Model Provider to Direct Competitor
OpenAI’s decision to launch a legal vertical under Ironclad co‑founder Jason Boehmig marks a shift from supplying foundation models to competing head‑on in legal AI tools. According to Legal IT Insider, this move “signals the company’s intention to compete directly for a larger share of the legal AI market.” Boehmig brings experience from building AI‑powered contract review and redlining on top of OpenAI models, helping set expectations for what generative tools can do in legal work. The timing follows Anthropic’s release of Claude for Legal and its partnerships with established providers such as Thomson Reuters and CoCounsel. OpenAI executives have stated that “the model alone is no longer the product,” and legal fits their push into agents and workflow automation because it mixes high‑value reasoning with document‑heavy processes. For firms, this means AI giants are no longer behind the scenes; they are vendors with their own roadmaps and pricing power.
Pressure on Startups and the Changing Vendor Landscape
As OpenAI, Anthropic and Microsoft each pursue dedicated legal offerings, the traditional legal tech ecosystem is under strain. Artificial Lawyer describes one scenario in which “Big Tech would enter the legal tech field at scale and eat everyone’s lunch,” with in‑house teams especially likely to move to the “Giant Three” for routine contract work and related tasks. Contract lifecycle management and other contract‑focused providers face a steep challenge: some will try to sell, some may fail, while data‑rich systems and tools that do not trade on productivity alone are somewhat more insulated. Law firms are unlikely to commit to a single large language model, since improvements from one provider can be accessed through various SaaS tools. Still, as more work consolidates onto enterprise platforms run by AI giants or large firms, smaller vendors risk being squeezed into narrow niches or acting as implementation partners rather than product leaders.
What This Means for Your Firm’s AI Strategy
For most firms, the question is no longer whether to invest in legal AI tools but how to structure that investment in a market dominated by a few powerful players. Large practices with strong balance sheets may follow Kirkland, building proprietary platforms on top of one or more general models and treating AI as core infrastructure. Mid‑size and smaller firms are more likely to standardize on offerings from OpenAI, Anthropic, Microsoft or legal SaaS vendors that integrate those models. In‑house teams, with fewer legacy vendor ties, may move fastest to all‑purpose AI platforms bundled across the business. Across all segments, expect pricing to reflect enterprise commitments rather than per‑seat experiments, feature development to favor deeply integrated workflows, and adoption timelines to shorten as legal technology investment becomes a board‑level topic rather than an innovation side project.