Kirkland & Ellis bets $500M on its own generative AI platform
What's the deal? Kirkland & EllisDealroom has a profile for this one. Try Dealroom →, the world's highest-grossing law firm, is spending $500M to build a proprietary generative AI platform rather than licensing off-the-shelf tools. Firm chair Jon Ballis told the Financial Times the firm expects to spend more than $100M this year alone, with hundreds of millions more over the next three to four years.
The platform is being developed with input from roughly 250 lawyers — including 100 partners — alongside 180 technology professionals. External tech companies are helping build it but are contractually barred from reselling the platform or its components to rivals.
The goal: an end-to-end system that can handle complex legal work from scoping through execution, rather than a patchwork of point solutions for document review, due diligence, and drafting.
Why now? Kirkland's logic is simple: if every firm can buy the same AI, it stops being a competitive advantage. "We don't get hired for the floor," Ballis said, arguing that widely available tools merely raise baseline standards across the industry.
The firm can afford the gamble. Kirkland recently became the first law firm to break the $10B revenue barrier, reporting roughly $10.6B in global revenue. Profits per equity partner hit a record $11.1M, up about 20% year-on-year. The $500M outlay represents roughly 1% of annual revenue.
What could go wrong? Building bespoke AI is expensive and risky. The spend will come out of current revenues, meaning a short-term dent in partner distributions — a sensitive matter even for partners earning eight-figure sums.
There is also no guarantee a law firm can out-engineer dedicated AI companies. Off-the-shelf tools from vendors like Harvey, backed by frontier labs such as AnthropicDealroom has a profile for this one. Try Dealroom → and Google, are improving fast. A custom platform that falls behind the commercial curve could become a costly white elephant.
The signal: Kirkland's $500M proprietary AI play is notable precisely because it runs counter to the prevailing legaltech trend, where even the largest firms have opted to license platforms from startups like Harvey rather than build from scratch. As a mature firm already clearing $10.6B in revenue, Kirkland has the financial headroom to treat AI as a capital project rather than a software subscription — a luxury few competitors can match. The bet will be watched closely across professional services as a test case for whether institutional knowledge, locked inside a bespoke system, can outperform the rapid iteration cycles of dedicated AI vendors.
Read more: lawfuel.com
