Kirkland & Ellis is preparing to spend more than $500 million building internal artificial intelligence technology as large law firms race to gain an edge in how legal work is handled, priced, and delivered to clients.
The firm is developing AI systems using the knowledge and working methods of its own lawyers rather than relying entirely on external software providers. Chair Jon Ballis said the goal was to deploy the “collective intelligence” of the firm throughout its operations. More than $100 million is expected to be spent this year alone, with additional investment planned over the next several years.
Firms across the legal industry are currently using many of the same AI providers, including Harvey and Legora. That may improve efficiency across the profession, but it also risks narrowing the gap between firms that have traditionally charged higher fees for specialist corporate work.
Ballis suggested Kirkland has little interest in competing with technology available to every rival firm. Widespread AI adoption, he said, was “raising the floor for everyone” in the legal market, but Kirkland & Ellis was not hired “for the floor”.
According to the Financial Times, roughly 180 technology professionals are involved in the project alongside Kirkland engineers and data scientists. The platform is also being developed using input from around 250 lawyers, including approximately 100 partners, with the aim of embedding legal knowledge directly into day-to-day operations.
Large firms are already signing deeper AI partnerships instead of simply licensing software. Earlier this year, Freshfields announced a collaboration with Anthropic focused on specialist legal AI products, while firms across the US and UK continue striking agreements with AI providers at pace.
Courts in both the US and UK are already dealing with cases involving lawyers who relied too heavily on generative AI systems. Sullivan & Cromwell recently acknowledged in a US bankruptcy matter that a filing contained multiple AI-generated hallucinations, while Pinsent Masons was reprimanded in London after false submissions linked to AI use appeared before the court.
Those incidents have increased scrutiny around lawyer supervision duties, verification obligations, and professional negligence exposure when AI-generated material enters filings or legal advice.
Ballis indicated that AI could speed up the move away from the traditional billable-hour model as firms automate more routine work previously handled by junior lawyers. Kirkland already handles some matters using value-based pricing, with clients paying for outcomes rather than lawyer time.
Changes in pricing structures could affect how firms recruit associates, allocate work, structure teams, and measure profitability. Firms with the financial capacity to build internal AI tools may also gain speed and operational advantages over rivals still dependent on external vendors.
Kirkland & Ellis reported more than $10.6 billion in revenue last year, giving the firm financial resources that many competitors cannot easily match. Spending at this scale is likely to increase pressure on firms that lack the capital to build internal AI capabilities while clients continue demanding faster legal delivery and lower costs.
AI is starting to become part of how firms control workflow, pricing, and client relationships rather than simply another office software tool.
Control over legal AI may increasingly determine which firms can deliver work faster, charge differently, and protect profit margins as clients demand more efficiency from outside counsel.