A Spotter’s Guide to the AI Law Firm

Under the selection pressure of legal AI, a new type of law firm has emerged: the AI-native law firm. The AI-native firm leverages technology to complete legal work, rather than associates in the traditional law firm structure.

Within the genus Lex Firma Intellegentiae Artificialis are four distinct species. This article is a field guide covering their distinguishing features, their favoured habitats and how to identify them in the wild.

Emerging first from the primordial soup is the full AI law firm. This is a technology-first product offering. Routine work is generated autonomously by the platform, with outputs overseen by qualified lawyers. Clients are charged per output.

Garfield was the first AI firm to be regulated in the UK by the Solicitors Regulation Authority. Garfield aims its services at those pursuing small claims and debt recovery.

Garfield looks and operates much like a technology product, so much so that the firm’s co-founder, Philip Young, refers to Garfield’s customers as service ‘users’ as opposed to clients. Users log on to the Garfield platform, upload documents, draft and send letters to advance a debt recovery matter. No interactions with a human lawyer.

There are also no office hours. “Garfield is available 24/7”, Young says, “if you want to start a small debt claim at three in the morning, you can.” Once a document is generated, and the user is comfortable with Garfield’s work, it is sent out and the user pays a flat fee. No billable hours.

The role of the human lawyer in the full-AI firm is to provide quality assurance for the work produced by the technology. Young operates as the oversight layer, reviewing AI-generated content, handling escalations and tackling any exceptional circumstances.

The net result is Garfield is more competitively priced than a traditional law firm, and faster too, Young argues. “Garfield is as useful for a sole trader with one outstanding debt who wants to have the process done properly at a fraction of the cost that a traditional law firm would do, all the way up to a really big company that has thousands of debts. The advantage of it is not just a saving in cost, but even more so the saving in time.”

The full AI firm is something radically different from the traditional law firm. At its core is the high degree of automation: the lawyer is the supervisor of the AI, not the primary producer of work. It is a service that behaves like software. It is an always-on law firm that is self-service and scales according to need, providing an economical option for those who have found legal services prohibitively expensive in the past.

The data-driven boutique takes a product-led approach, using AI tools to turn legal work into structured data and intelligence. Outputs can be used by in-house teams or used in conjunction with the firm’s lawyers on strategic matters.

Covenant, founded in 2024 by Jen Berrent and Richard Perris, focuses tightly on private market investors, assisting in portfolio management and handling transactions.

The productised side of Covenant is in the structured extraction and analysis of investors’ data, as Jen Berrent explains: “Clients send us investment documents […] we then take that document and have a survey or set of questions that we query against it, about 150 questions, creating a schema.” That schema is checked by the firm’s AI models. “It's a very fast review,” says Berrent, but when the AI flags an issue, “lawyers can open the document and make a judgment call.”

Covenant differs in what comes next. The client, with their analysis in hand, can either take that information away or engage Covenant’s lawyers as deal counsel. “Think of it like tech-enabled services,” Berrent says; “most of it is the technology, with a layer of human expertise on top.”

Clients pay a fixed fee for the review and analysis, then the firm charges by the hour for non-automated work. Covenant treats the billable hour as one tool among several, not the default. Berrent again: “Law firms just became beholden to the billable hour. We see it as a supplement.”

The data-driven boutique is built for clients like private investors: consumers of legal services whose needs do not always justify paying for premium associate time. Firms like Covenant commoditise legal analysis, while retaining lawyers for strategic decisions, negotiations and judgement calls where their expertise adds the most value.

The senior-led fractional firm moves from product-led services into the realm of the cyborg lawyer: practitioners who use legal AI to carry out tasks.

Three Points Law, founded by Mishcon de Reya alumni Simon Leaf and Tom Murray, advises start-up technology businesses and other high-growth companies on commercial contracts, technology transactions and day-to-day legal operations. Pricing is by fixed monthly retainer.

A small number of senior lawyers handle everything, making heavy use of legal AI systems like Legora to do so. The result, Leaf says, is that a single Three Points lawyer can handle matters “that would previously need a team of three or four at a big firm, or one or two in-house.”

Differing from an autonomous AI approach, Three Points’ team members act in concert with the AI tools, with strategic advice remaining the human’s domain. “[AI] tools get you about 65–75% of the way there,” Leaf says, “but they don't get it 100% right and often lack client-specific context. The human experience is crucial to pick up the remaining 25% — the context, sensitivities, or nuances that machines miss.”

The result is a model that preserves the close client relationship and commercial judgement associated with in-house counsel, while using AI to deliver those services more efficiently and flexibly than traditional firms.

The final breed aims to offer the quality and range of BigLaw, but with a leaner internal model.

Pierson Ferdinand debuted in 2024 with 130 lawyers offering a full-service commercial practice. Unlike other firms of this size and scope, Pierson Ferdinand does not employ associates. All work is carried out by partners assisted by AI tools.

Michael Pierson, one of the firm’s founders, emphasises that, from the client’s perspective, the experience is the same as they might expect from an AmLaw100 firm. Clients can expect to benefit from “high execution quality and lots of personalised attention.”

Legal AI in this case functions in place of the associate. Pierson explains AI delivers “first drafts, summaries, due diligence, indexing… all those things that we as junior associates used to do and charge our clients for.”

Pierson Ferdinand’s charging model is less radical. “Most clients still are okay with hourly billing,” Pierson says, “given the fact that our rates are lower.” The firm’s model does allow for more flexible fee structures, though. Pierson says: “We entertain alternative fee arrangements all the time […] AI is pushing us in the direction to facilitate the delivery of outcomes for the client, less than physical time that's being billed.”

Not every AI-powered partnership leverages technology in the same way. Norm Law is the captive firm of AI software company Norm AI. Like Pierson Ferdinand, it targets the work traditionally performed by large commercial firms. Unlike its cousin, however, Norm's objective is not to remove the associate layer, but to recreate it in software.

Mike Schmidtberger, Norm Law’s chairman and former chair of Sidley Austin, explains the typical client journey: “The in-house department sends an email. It goes directly to an AI agent. The agent then does the assignment in the first instance. […] The senior associate will then improve the work, or certainly review it, and […] either send it directly to the other side for negotiation or send it to the in-house legal department for their review.”

Sitting above the entire process is a Norm partner who holds ultimate responsibility for the deal. Norm only charges per-deal, not by the hour.

Where Pierson Ferdinand dispenses with associates entirely, Norm retains them alongside the AI agents. Why not remove associates altogether? Schmidtberger invites us to think of Looney Tunes. Having partners handle every task is like Elmer Fudd going hunting with an elephant gun: “you blow up the duck if you shoot it with an elephant gun.” In the same way, Schmidtberger argues partner time is best spent on the most strategic matters: “you're losing all the efficiency of the AI if there's nobody in between to manage the interim.”

Norm’s ambition is to accumulate and retain institutional knowledge in software rather than people. Schmidtberger says, “We are trying to make sure that our autonomous associates as well as our humans continue to learn and partner with the client.” Norm's bet is that by embedding client knowledge in its agents, it is made permanent: "Our proposition isn't that partners are valuable because of the work they pore over, or because they have a relationship they could port. Their value is their service to the client, tied inextricably to technology."

Law is currently undergoing its own Cambrian explosion, with new species of AI-native legal firms rapidly emerging. What does this mean for the traditional law firm? Extinction seems unlikely. Evolution, however, is already underway.

The clearest evidence comes from the industry's apex predators. Kirkland & Ellis is investing USD500 million into its own AI platform. Legal AI is not replacing the legal profession; it is changing the environment in which it operates. The firms that thrive will be those best able to adapt.

Chambers’ rankings make it easy to spot the best, most innovative UK law firms.

View rankings