© 2026 JSH Law Ltd. AI-assisted original artwork created for JSH Law using supplied JSH Law branding and portrait photography. All rights reserved.Legal AI Is Rebuilding the Legal Services Market — But Who Will It Serve?
Legal AI Is Rebuilding the Legal Services Market — But Who Will It Serve?
Legal technology is entering a new phase. The most significant developments are no longer simply about whether lawyers can use artificial intelligence to research faster, summarise documents or produce first drafts. They are about the infrastructure through which legal work is received, analysed, performed, supervised, priced and delivered.
For several years, much of the discussion around artificial intelligence in law has focused on productivity.
Can AI help lawyers draft more quickly? Can it review large volumes of documents? Can it identify relevant authorities? Can it reduce the time spent on repetitive legal work?
Those questions remain important, but they are no longer sufficient.
Recent developments across the legal technology market suggest that artificial intelligence is moving beyond isolated tools and into the architecture of legal services themselves.
On 4 June 2026, Kirkland & Ellis and Palantir announced an exclusive AI-powered enterprise platform intended to transform private equity fundraising.
At the other end of the market, Lavern is testing whether sophisticated multi-agent legal analysis can be made accessible, adaptable and open source.
Meanwhile, Wordsmith has raised $70 million to develop technology through which in-house legal teams can manage more of their work internally, while ATJustice is using AI to help members of the public identify suitable lawyers rather than attempting to replace them.
These may appear to be very different products serving very different users. Taken together, however, they reveal a common direction of travel.
Legal AI is moving from the margins of legal work into the infrastructure of legal services.
Four Legal-Tech Developments, Four Different Futures
The current market is not moving in one direction. It is dividing into several competing models, each of which raises different questions about professional value, legal service delivery and access to justice.
| Development | Primary User | What It Suggests | Strategic Question |
|---|---|---|---|
| Kirkland & Ellis and Palantir | Elite private practice | Large firms are building proprietary systems around their own institutional knowledge and workflows. | Will legal knowledge become a protected technology asset? |
| Lavern | Developers, lawyers and smaller organisations | Open-source multi-agent systems may make sophisticated legal analysis more widely available. | Can transparency and accessibility compete with proprietary scale? |
| Wordsmith | In-house legal teams | Corporate legal departments are becoming technology-enabled operating platforms. | How much routine legal work will remain with external advisers? |
| ATJustice | Members of the public | AI may be most valuable when it helps people reach the right human professional. | Can technology improve access without creating false confidence? |
From General-Purpose Tools to Proprietary Legal Infrastructure
Kirkland and Palantir’s new platform is designed to support private equity fundraising across the lifecycle of fund formation work. According to the firm, it will help scale institutional knowledge and judgment, streamline complex workflows and support clients and investors more efficiently.
That wording is significant.
The proposition is not simply that an AI system can draft a document more quickly. The proposition is that a law firm can capture, structure and deploy aspects of its accumulated knowledge across a complex area of practice.
For large firms, this creates a new form of competitive advantage.
Legal knowledge has always been embedded in people, precedents, supervision structures, internal know-how and professional relationships. Proprietary AI systems may allow firms to operationalise that knowledge more consistently and at greater scale.
The future legal-tech divide may not be between firms that use AI and firms that do not.
It may be between:
- firms that rely on broadly available tools;
- firms that configure those tools around their own workflows;
- and firms that build proprietary systems around their data, expertise and commercial strategy.
A tool available to every competitor is unlikely, by itself, to create lasting differentiation.
The more valuable question is whether a legal organisation can identify which parts of its expertise should be standardised, which parts should remain dependent on individual judgment, and how technology can support both without confusing one for the other.
The Rise of the Agentic Legal System
Lavern represents a very different approach.
Rather than presenting a single conversational interface that produces a single answer, Lavern has been described as an open-source multi-agent legal system. Its specialist agents can examine different aspects of a legal problem, challenge one another’s reasoning and contribute to a structured output.
This reflects a broader movement towards “agentic” AI: systems intended to perform a sequence of tasks, use different tools or perspectives, and work towards a defined objective with a degree of operational autonomy.
In legal services, that could eventually mean systems that do more than answer questions.
They may identify issues, organise evidence, compare documents, test arguments, locate inconsistencies, prepare preliminary chronologies or flag matters requiring human review.
There is obvious potential in that model. There is also obvious risk.
A multi-agent system may create the appearance of deliberation, challenge and consensus. That does not mean it has exercised legal judgment, understood the human context of a case or produced a reliable conclusion.
An AI system may simulate a legal team without carrying the duties, accountability, experience or ethical responsibility of one.
That does not make such systems useless. It makes their design, supervision and deployment critically important.
In-House Legal Teams Are Becoming Technology Platforms
Wordsmith’s recent funding announcement illustrates another important shift.
The company is building a platform through which in-house legal teams can receive requests, handle routine work, prepare documents and present matters for legal review. Its stated aim is to help businesses keep more legal work in-house and reduce unnecessary external counsel spend.
For private practice, that is not merely a productivity story. It is a market-structure story.
When legal departments become better able to triage, standardise and resolve routine work internally, they may instruct external advisers later, less frequently and with a more tightly defined brief.
External lawyers may therefore be asked to provide less process and more judgment.
That could increase the value placed on specialist expertise, strategic insight, advocacy, complex negotiation and the ability to address genuinely difficult questions.
It may also reduce the commercial viability of work that has traditionally depended on volume, opacity or inefficient internal processes.
The firms most likely to succeed will not be those that defend every existing task as inherently legal or inherently human. They will be those that can articulate where their professional value truly lies.
Access to Justice Requires More Than Automated Answers
The development that interests me most is ATJustice.
ATJustice uses AI to help people describe their legal problem and identify lawyers who may be suitable to assist them. Its founders have emphasised that AI should improve access to legal services rather than replace licensed professionals.
That distinction deserves more attention.
Much of the access-to-justice debate assumes that the central problem is a lack of legal information. In some circumstances, that is correct. Clear explanations, guided forms and better public-facing resources can make a substantial difference.
But many people do not simply need information.
They need help to understand what kind of problem they have, what support is available, what must be done urgently, what evidence matters, what procedure applies and when professional representation is necessary.
In family justice, for example, a litigant in person may have access to court forms, procedural rules and general guidance while still being unable to identify the real issues in dispute or present relevant evidence effectively.
The best use of AI may therefore be neither to replace lawyers nor to provide the public with confident-sounding legal answers.
It may be to create better routes into appropriate human support.
Responsible access-to-justice technology could assist with:
- early identification of urgency or safeguarding concerns;
- clearer referral to regulated legal advice or specialist support;
- better preparation before a legal appointment;
- more effective triage of documents and procedural questions;
- and improved signposting for those who cannot afford traditional representation.
Used responsibly, technology can reduce the friction that prevents people from obtaining help.
Used carelessly, it can create a false impression that help has already been provided.
The Professional Question Is Not Whether AI Is Impressive
The pace of legal-tech development creates a temptation to assess products primarily by what they can produce.
Can the system draft a memorandum? Can it review a contract? Can it identify relevant authorities? Can it prepare a chronology? Can it answer a legal question in seconds?
Those are useful demonstrations, but they are not the complete test.
For legal professionals, the more important questions include:
- What information was provided to the system?
- What information was excluded?
- Can the output be verified?
- Can the reasoning be interrogated?
- How is confidential or privileged information protected?
- Who is responsible for errors?
- What degree of human supervision is required?
- Does the system create or reinforce bias?
- Can the user distinguish between legal information, legal analysis and legal advice?
- Will the technology improve procedural fairness, or merely increase the speed at which one party can produce material?
These questions are particularly important where AI-generated work may affect litigation, judicial decision-making, vulnerable parties or individuals without legal representation.
The legal profession cannot outsource its duties to a system, however capable that system appears to be.
Nor should responsible caution be confused with resistance to innovation.
The most credible legal-tech strategies will be those that combine technical ambition with professional discipline.
What This Means for Lawyers, Firms and Courts
1. Legal knowledge is becoming an operational asset
Firms and legal teams should consider how their knowledge is captured, maintained, supervised and made available. A disorganised bank of precedents is not an AI strategy. Neither is a collection of individual subscriptions purchased without governance or integration.
2. Workflow design matters as much as model capability
The most valuable legal systems may not be those with the most impressive standalone answers. They may be those that fit carefully into the way legal work is received, reviewed, escalated and completed.
3. Human judgment will become more visible, not less important
As routine work becomes easier to automate, clients and employers will place greater value on the ability to identify the real issue, assess risk, exercise judgment, communicate clearly and take responsibility for a course of action.
4. Access to justice must be designed into legal technology
There is a danger that the most capable systems will be developed exclusively for the organisations already best resourced to obtain legal assistance.
Innovation should also address the needs of individuals, litigants in person, small practices, advice organisations and public institutions.
5. Courts will increasingly encounter AI-shaped material
Even where judges do not use a particular technology, they may be asked to determine cases in which documents, arguments, evidence summaries or authorities have been produced or influenced by AI.
Reliability, transparency and procedural fairness will become increasingly important.
The New Legal-Tech Divide
There is a real risk that legal AI could widen existing inequalities.
Elite firms may have proprietary systems built around deep institutional knowledge. Corporate legal teams may use technology to reduce external spend and increase control. Well-resourced parties may be able to analyse large volumes of material more quickly and present more polished cases.
Meanwhile, individuals without representation may still struggle to understand a court order, identify relevant evidence or find suitable legal assistance.
That outcome is not inevitable.
Open-source projects, responsible referral platforms, better public legal education and carefully designed support tools could make legal help more accessible and more effective.
But access to justice will not improve simply because more technology exists.
The measure of legal innovation is not only whether it makes legal work faster. It is whether it makes the legal system more understandable, more accountable and more capable of delivering justice.
Conclusion
The legal AI market is no longer defined by isolated drafting assistants or experimental chatbots.
It is becoming a contest over legal infrastructure, institutional knowledge, workflow ownership, professional judgment and access to services.
Kirkland’s proprietary platform, Lavern’s open-source multi-agent model, Wordsmith’s in-house legal operating system and ATJustice’s referral technology each point towards a different possible future.
The profession should be ambitious about what legal technology can achieve.
It should also be precise about the problems it is trying to solve.
My own interest lies in the point at which legal technology, litigation support and access to justice meet: how better systems can help legal professionals work more effectively while also improving the experience of those who must navigate legal processes without conventional representation.
That is where some of the most difficult questions remain — and where some of the most valuable legal innovation is still waiting to be built.
Legal Technology, Family Justice and Access to Justice
I am particularly interested in conversations with legal professionals, legal-tech organisations, courts, researchers and access-to-justice projects exploring how technology can improve legal service delivery without weakening professional accountability or procedural fairness.
There is significant scope for better systems to support evidence analysis, litigation preparation, legal triage, public legal education and the experience of litigants in person.
Thoughtful legal innovation should not begin with the question, “What can the technology produce?” It should begin with the question, “What problem are we trying to solve, and for whom?”
Sources and Further Reading
Need Support With Family Court Preparation?
JSH Law provides structured litigation support for litigants in person, including assistance with case analysis, chronologies, witness statements, evidence organisation, hearing preparation and procedural guidance.
Book a 15-minute initial consultation to discuss the support you may need.
Regulatory & Editorial Notice
This article is provided for general information and commentary only. It does not constitute legal advice and should not be relied upon as a substitute for advice from a suitably qualified legal professional.
JSH Law is not a firm of solicitors and is not regulated by the Solicitors Regulation Authority. JSH Law provides litigation support and McKenzie Friend services for litigants in person and does not undertake reserved legal activities.
References to third-party organisations, products, publications and services are included for editorial and public-interest commentary. JSH Law is not affiliated with, endorsed by or responsible for any third-party product or service referred to in this article unless expressly stated otherwise.
© 2026 JSH Law Ltd. All rights reserved.
© 2026 JSH Law Ltd. All rights reserved.
© 2026 JSH Law Ltd. All rights reserved.
Leave a Reply
Want to join the discussion?Feel free to contribute!