What will the law firm of the future actually look like?
My interest in this question did not arise only after the recent AI boom. Since graduation, I have essentially been working at the intersection of law and the internet. In the early years I worked on legal internet products, then on business growth for lawyers, then on digital tools for lawyers and law firms, and now at Mankun Law Firm I provide legal services related to Web3, AI, and fintech. Many things looked promising back then, but only when you actually get to the front line of the business do you realize there is a very subtle tension between the legal industry and the internet industry.
The internet favors standardization, scale, and datafication; it likes putting supply and demand on a single platform so transactions become faster, cheaper, and more transparent. But legal services happen to be a business that depends heavily on trust, experience, judgment, and boundaries of responsibility. Clients are not buying a standardized commodity, and lawyers are not simply putting their time on a platform waiting for orders. There is too much human judgment, emotion, risk, and distribution of interests involved.
So I am now increasingly certain of one thing: if the law firm of the future is really going to emerge in China's legal industry, it most likely will not evolve naturally out of the old model of traditional law firms, nor will the transformation be accomplished simply by purchasing an AI system. It has to grow all over again in new client relationships, new collaboration platforms, new service models, and new industry soil.
This is not a slogan; it is more like a judgment left behind after years of trial and error. In this article, I am sharing some personal reflections and opening a discussion with everyone.
The platform model is very tempting
What first excited people about the legal internet was the vision of the platform economy.
In 2013, what we wanted to do was straightforward: clients on one end, lawyers on the other, with a platform in the middle to improve connection efficiency. Clients would post their legal needs online; lawyers who saw them could submit service proposals and quotes to the client. The platform would then use publicly available data to help clients better judge whether a lawyer was suitable.
On paper, this logic holds. In the past, clients found lawyers largely through acquaintances; information was opaque and the cost of choosing was high. If the platform could analyze a lawyer's past cases, location, areas of expertise, and win rates, clients would no longer have to rely entirely on a friend's word. Lawyers would also gain an additional online channel for client acquisition, which seemed fairer, especially for young lawyers without much offline resources.
At that time, we did seriously consider using China's publicly available court judgments to analyze lawyers' capabilities. For example, how many cases a lawyer had handled in a particular region or case type, what the outcomes generally looked like, and whether one could discern something about their professional ability and case experience. In this way, the long-standing information asymmetry in the legal industry seemed to have a chance of being broken.
But once we actually got to work, we found it was not as simple as we had imagined.
On the one hand, for certain well-known reasons, the publication of court judgments itself is affected by policy, the scope of disclosure, and data completeness. At certain points in time, the pace and availability of published judgments changed, making it very difficult for a platform to rely on this data source for long-term, stable, and objective lawyer evaluation. If the data is incomplete, no matter how beautiful the visualization or the report analysis is, it is useless.
On the other hand, the platform will also expect lawyers to supplement information themselves. The problem is, no lawyer is particularly proactive about putting online the cases they handled poorly, lost, or would rather not have others see. What lawyers are willing to showcase is usually the side that proves their professionalism and success. Clients want complete information, platforms want to form objective data, and lawyers want to present their strengths; put these three things together, and they do not automatically add up to genuine transparency.
At this point you realize that legal services are not simple information matching. For a platform to turn lawyers' capabilities into data, it first needs data that is sufficiently stable, complete, and credible; yet in the real industry, the sources, boundaries, and motivations behind the data are all unstable.
The bigger problem lies in the business model. A platform's value comes from reducing transaction costs, but the platform also needs revenue. In the legal industry, a typical sales commission or platform service fee runs to roughly the 20% range. For a young lawyer, this is not an abstract percentage but a very tangible reduction in income. Clients do the math, and so do lawyers. Once the two sides have established contact, the incentive to bypass the platform emerges.
Many people call this "order skipping." From the platform's perspective, this is of course a problem; but if you stand in the shoes of the lawyer and the client, you will find that they have their own economic rationality. The lawyer wants to preserve income, the client wants to lower costs, and the platform wants to collect service fees; each side makes sense on its own, but combined, the structure is very fragile.
So the platform model sounds very internet-native and very elegant, but applied to legal services, it easily gets stuck in two places: first, objective information is very hard to truly form, and second, the distribution of benefits is very hard for all participants to accept over the long term.
Exploring Business Growth
After the platform model failed to run smoothly, we also took a step back and began thinking about another question: instead of building a large, all-encompassing legal platform, could we focus specifically on helping lawyers grow their business?
I began experimenting more in this direction in 2019. Put simply, it meant organizing potential client needs and referring them to suitable lawyers. This covered both relatively basic consultation needs and legal service needs that clients had already expressed fairly clearly.
On a small scale, this can work. For example, if you have worked with certain lawyers over a long period, you know their case types, service capabilities, pricing habits, and client communication styles, and you also know roughly how to assess and convert a certain type of client. At that point, business growth is not simply "selling leads"; it is more like making a more accurate match between clients and lawyers.
But once you try to turn it into a scaled platform business, the problems quickly start to change in nature.
Many people underestimate how difficult it is for lawyers to take on clients online. When clients come to consult, they often do not arrive with a clear-cut legal question. They may come with anxiety, with misunderstandings, with budget constraints; they may have concealed part of the facts, or they may not even have figured out for themselves what problem they actually need to solve. The lawyer must first understand what the client truly wants, then assess whether that need can be turned into a service, and then explain the risks, build trust, negotiate pricing, and define the boundaries of deliverables.
Not every lawyer is good at this whole process. Meeting clients offline and taking consultations online are not the same thing. With offline referrals and endorsements from friends, a foundation of trust already exists between the two sides, and many things become easier to communicate; this is why everyone feels that clients referred offline are more likely to reach an engagement. But online, that foundation of trust does not exist. Online clients are faster-moving, more scattered, more prone to comparison, and more likely to switch back and forth among different lawyers. If a lawyer lacks sufficiently strong communication skills, product awareness, and conversion awareness, the client's consultation may well go without being truly satisfied.
There is also a more practical problem: the unit of measurement is unclear.
Does one consultation count as a valid lead? When a client moves from a vague need to a clearly defined one, who did the educating in between? If a lawyer spends a great deal of time communicating but the client never signs on, who bears that cost? And if the client eventually does sign on, should the fee be charged per lead, as a percentage of the case, or per service item? As long as these questions cannot be answered clearly, so-called business growth can easily degenerate into a short-term traffic grab.
That is why I have become increasingly cautious about this kind of model. It is not entirely unworkable, but it is very hard for it to become the underlying logic of the law firm of the future. Because what it solves is only a small piece of the client acquisition problem; it does not solve the firm's own organizational capability, service capability, or long-term trust.
Digital tools for the legal profession
Later on, another wave of lawyer tools emerged in the industry.
Case management, contract management, document automation, client management, knowledge bases, project management—people were building all of these. By the AI era, many of the names had been swapped out again, becoming intelligent search, intelligent drafting, lawyer agents, and AI assistants.
I am not opposed to tools. Quite the opposite—I am myself a heavy tool user. That Mankun has been able to run this far over the years is also inseparable from digital tools. The real problem is that, for the digitalization of China's legal profession, the difficulty has never been the tools themselves, but the organization itself.
The SaaS business model is not easy in many industries in China, and the legal industry is an especially typical case. It often does not die at the first sale, but at renewal. When clients buy, they think it is useful; a few months later they discover the organization has never truly adopted it, the processes have not changed, responsibilities have not shifted, and everyone is still working in their own way. Renewal naturally becomes very difficult.
There is another aspect unique to the legal profession: tools for individual lawyers and tools for law firm management are often two entirely different things.
A while ago, at Shanghai Hongqiao Airport, I saw DingTalk's advertisement for its "Wukong Lawyer Edition." I even posted a moment on WeChat reflecting that, after all these years of building digitalization software, Alibaba's product managers still have not truly realized that digital tools aimed at individual lawyers and digital tools aimed at law firm organizations are two different problems.
The vast majority of Chinese lawyers are relatively independent in their practice. Many lawyers use tools to record their own cases, manage their own clients, draft their own documents, and build their own knowledge repositories; these may have little to do with firm-level management. If you sell to individual lawyers, you have to educate, sell to, and deliver for one lawyer at a time; if you sell to law firms, you find that many firms lack the organizational capability to actually roll the tool out.
From the outside, the number of lawyers looks large and the market looks big. Once you actually get into it, you find that your sales costs and delivery costs are not calculated per law firm, but per individual lawyer. The user count looks large, but the education costs and delivery costs rise along with it, which is extremely demanding for a startup.
This is also why I remain fairly sober-minded when looking at AI legal tools today.
AI will certainly change how efficiently lawyers work, but it will not necessarily bring about a qualitative transformation at the industry level on its own. Lawyers who were willing to use digital tools in the past will now use AI more actively; teams that were willing to move collaboration online and to deposit knowledge and projects into systems will likewise keep moving forward. Yet from the organizational standpoint of many law firms, the problems will not change that quickly: there is no willingness, no motivation, and more importantly, no capability to drive deep, organization-wide transformation.
Tools will make the advanced even more advanced, but they will not automatically turn a loosely structured organization into a modern one.
What the legal profession truly lacks is a way of working that allows clients, projects, knowledge and responsibility to accumulate over the long term.
Mankun Is the Experimental Ground for the Law Firm of the Future
When we started Mankun in 2021, we in fact began with precisely this judgment in mind.
Many friends came to know Mankun because over the past few years we handled numerous cases and did extensive industry promotion and education in Web3, blockchain and cryptocurrency. That label is certainly accurate. Over the past few years, Mankun has indeed had a significant presence in cryptocurrency-related work, to the point that many people have overlooked what we do in AI and fintech.
But to understand Mankun solely through its business labels is to see only one side of it.
From the very beginning, what concerned me more was another question: if someone who came out of legal technology and the legal internet were to build a law firm from scratch, could they rebuild the law firm according to the logic of the next-generation internet and the digital era?
As the venture developed over the past few years, Mankun's core business gradually expanded into three areas: Web3, that is, the new-generation internet business built around blockchain and crypto assets; AI; and fintech. On the surface, these are three business segments. When actually serving clients, however, one finds that the needs behind them are highly similar.
Clients are going global, confronting the regulatory environments of different regions, and dealing with company formation, cross-border data transfer, asset issuance, payment and settlement, license applications, contract design, dispute resolution and compliance boundaries. Web3 clients and AI clients appear to be in different industries, but they all face issues of global operations, explaining technical products, compliance of business models and multi-jurisdictional coordination. With the current flourishing of AI, the determination of AI identity, the protection of data and privacy, and especially monetary settlement between AI agents have brought to the fore the infrastructure questions of next-generation fintech — and these are the very problems we help our clients solve at every moment.
Clients of this kind are inherently online, global and digital. Every day they switch between Feishu, WeChat, Telegram, Tencent Meeting, Zoom, Google Docs, on-chain explorers, trading platform backends and various product systems. Their business already flows online. If lawyers still deliver services relying only on paper-based processes, offline verbal handoffs, personal file folders and scattered WeChat groups, misalignment is only a matter of time.
So Mankun was not created to chase a new-economy label. For me, it is more like an experiment in the future form of a law firm: whether new client groups, new ways of collaborating, new brand expression, and new service processes can be used to rethink the legal services organization.
Clients Will Identify You Online
The first change future law firms must confront is that the way clients come to know and identify lawyers has already changed.
In the past, referrals from acquaintances were crucial to finding a lawyer. That will not disappear. Legal services are high-trust services, especially when large sums, high risks, and complex relationships are involved; clients will still ask friends, peers, investors, and former clients.
But the change now is that even if a client first hears about you through an acquaintance, they will verify you again online. They will search your name, your firm, what you have written, which industries you have worked in, and whether you have consistently expressed views on a particular issue. They may read WeChat official accounts, or they may watch Channels, Douyin, or Xiaohongshu, or they may simply ask an AI search tool.
This is no longer “marketing” in the traditional sense. It is more like advance verification of a trust relationship.
Before a client actually contacts a lawyer, they may already have read several public answers, asked several AI tools, organized a round of materials, and even formed their own preliminary view. One figure in Clio’s 2025 legal trends report is worth noting: more than 50% of clients first turn to AI for preliminary information on legal issues. That figure cannot simply be transplanted to the Chinese market, but the direction is clear: clients no longer come to lawyers empty-handed.
This raises the bar for lawyers. Clients will not only ask, “How do you view this legal issue?” They will also assess whether you truly understand their industry, whether you can explain complex problems clearly, whether you have produced sustained output in the past, whether your expression is consistent, and whether your judgment has boundaries.
So I have always believed that no lawyer in the future can avoid online expression. Not every lawyer needs to become an influencer, and not everyone needs to shoot short videos every day, but at a minimum they must meet a passing standard of professional credibility online. The clients you serve, the industries you understand, the problems you can solve, and whether you have a continuous record of judgment—these things cannot remain hidden only in offline word of mouth.
In the past, trust in legal services was transmitted mainly through chains of acquaintances. That chain will still exist in the future, but it will be overlaid with a layer of public content and digital traces. Acquaintances bring the client to your door; online content determines whether the client is willing to keep going.
Digital Collaboration Inside the Law Firm
Looking further inward, collaboration within a law firm must be digitalized.
Meeting in person is, of course, important. In fact, I believe that the further we move into the AI era, the more precious it becomes for people to sit down and have a genuine conversation. Much trust, judgment, and emotion cannot be fully resolved through online tools alone.
But the vast majority of a law firm's day-to-day collaboration should no longer rely on fragmented messages in WeChat groups, the memory of a particular partner, folders on someone's personal computer, or a question like "where is that version you sent me last time." As client projects multiply and teams grow more complex, a collaboration model propped up by memory and verbal handoffs will continuously drain people.
From its very founding, Mankun chose Feishu as its underlying collaboration platform. To this day, I still believe Feishu is one of the best collaboration foundations for legal service teams in China. Many friends ask me whether a law firm should buy an OA system, adopt an industry SaaS product, or use a dedicated case management system. My answer has always been fairly direct: in many cases, there is no need to start by purchasing something that looks like a "law firm system."
If a law firm truly understands its own business processes, it can build many systems suited to itself on top of a general-purpose collaboration platform like Feishu, combined with low-code tools and AI coding. Client management, project advancement, contract workflows, knowledge bases, content collaboration, event operations, financial statistics, and risk alerts can all gradually take shape.
The key here is not the name of the software. The key is whether the organization is willing to distill and codify its own ways of working.
The problem with much industry software is that it forces you into its preset workflows. But the digitalization truly suited to a law firm of the future should turn its own experience, processes, judgments, and boundaries of responsibility into things that are traceable, reusable, and transferable.
Over the past five years, a large share of Mankun's internal work and the firm's overall operations have been carried out on Feishu. In 2022, Feishu described Mankun as "a law firm grown on Feishu," and Mankun's founding story can still be found on Feishu's official website today. I have always liked that description. It does not merely mean that we use a piece of office software; it means that our collaboration, projects, and knowledge have genuinely grown on a digital foundation.
The internal collaboration of the law firm of the future should not be "someone is very capable, so everyone goes to ask him." It should gradually become an organizational capability: what stage a project has reached, where the materials are, what the client has said, what the team has concluded, who is responsible for the next step, and whether risk points have been recorded—all of this should be visible, inheritable, and reviewable.
Client Service: Human-AI Collaboration
Client service itself will also be reshaped.
I find it hard to imagine that by 2026 there will still be law firms whose engagement contracts with clients need to be signed and completed on paper—yet this is indeed the current reality for the vast majority of law firms in China today. It is easy for us to envision the standard motions of lawyers serving clients in the future: signing completed online, meetings held online, documents collaborated on online—none of this is novel anymore. The bigger change is that AI will enter the client service process and become a very routine form of human-machine collaboration.
Recently, the more deeply I use AI, the more often I picture a concrete vision of the future: in the future, every client will have a "dedicated" AI assistant.
This AI assistant is not a generic public chatbot hastily built by some outsourced vendor to fob off clients, nor is it a customer-service bot that anyone can freely chat with. It should be a dedicated service gateway formed around a specific client, a specific project, or a specific matter. It understands the client's situation, knows which stage the project has reached, knows which materials have already been submitted, which issues have been discussed before, and which conclusions have already been confirmed by the lawyers. Behind it, it connects to the firm's knowledge base, the legal team's experience, and the project's memory, while still maintaining privacy segregation and access control.
When a client wants to confirm a process in the middle of the night, they do not necessarily have to wait for the lawyer to wake up. When a client wants to find a document discussed earlier, they do not necessarily have to scroll through a WeChat group for ages. When a client wants to know what to prepare next, the AI assistant can first provide structured prompts. When the lawyer is in court, resting, or driving, the client's basic needs will not go entirely unanswered. I have previously observed that more than half of clients' biggest complaints about legal services are that they often cannot reach their lawyer.
When it comes to matters requiring a lawyer's signature, confirmation of legal responsibility, major judgments, participation in negotiations, or emotional support for the client, the human lawyer steps in.
This may sound very forward-looking, but it is actually consistent with how the vast majority of lawyers in small teams serve their clients today.
Take Mankun Law Firm as an example. In all of our client service groups, there are currently lead lawyers, assisting lawyers, and paralegals. The direction we will explore in the future is this: certain information organization, process reminders, basic responses, and document retrieval—and even clients' basic Q&A—will be handled by the AI assistant. What clients truly care about is not who replies to the message first, but whether the service has continuity, whether the answers have a basis, and whether someone is accountable at the critical junctures.
Of course, legal AI should not be mythologized. Even professional legal research products can still produce inaccurate or fabricated information; simply handing a tool to its users does not automatically bring efficiency gains. Training, processes, and verification mechanisms all directly affect the results, and law firms in the future certainly will not—and cannot—simply hand clients over to AI. A more prudent approach is to embed AI within a verifiable, accountable, and traceable service system. AI can take on more of the foundational work, but the responsibility, judgment, and emotional support in legal services must still be borne by people.
This is only a matter of time.
In current industry discussions about AI legal services, many people talk about distilling lawyers' capabilities, lawyer skills, and legal agents. These are certainly all worth exploring, but I see them more as interim solutions.
AI has only just begun to enter work scenarios. We may not have even scratched the surface of its true application scenarios.
This is a bit like imagining the future of the internet before 1999. Back then, many people could hardly imagine that today, hailing a ride, ordering takeout, buying tickets, holding meetings, making payments, working, and consuming content could almost all be connected to online systems. People in the 1990s also found it hard to imagine how the mobile internet would later become embedded in real society.
Our imagination of AI today is likewise constrained by past internet experience. We tend to picture AI as a smarter chat box, a faster writing tool, or a cheaper assistant, but in the future it may rewrite many workflows themselves.
Within law firms, over the next five to ten years—or perhaps even sooner—the firm's operating model, client service, business development, team collaboration, and long-term client support may all be reorganized. Not every change will happen immediately, and not every tool will endure, but the direction is already clear: legal services will become increasingly online and increasingly systematized, and lawyers' individual capabilities will increasingly be amplified through tools and organizations.
At this point, the question to ask is no longer "should I use a particular AI tool," but rather: are my law firm, my team, and my individual capabilities ready to enter that form of service delivery?
An Old Map Cannot Show New Directions
From time to time, younger lawyer friends come to me to discuss career development paths in the legal profession, hoping I can offer some advice. But to be honest, I cannot give advice, because all innovation and development in human history rests on a very simple logic: do not fixate solely on the path your predecessors tell you to take.
Much of what seniors, partners, and mentors say is of course correct. Build a solid professional foundation, maintain client relationships, handle cases diligently, express yourself reliably, and be a trustworthy person—these principles will never go out of date. The problem is that while the principles may be right, the ways and directions of doing things may already have changed.
The traditional growth path for lawyers was often to put in the years, follow a mentor, build case sources, join professional circles, and accumulate clients. This path still has value, but it is no longer the only road. Young lawyers of the future may need to identify a niche industry earlier, continuously produce professional content, understand technology and business, participate in global projects, and leverage AI and digital collaboration to expand their service reach to a degree that a traditional solo lawyer could hardly achieve.
When an industry undergoes a qualitative transformation, the greatest danger is not slowness, but holding an old map while searching for a new road.
I am not saying that young lawyers should not learn from their predecessors. Quite the opposite: fundamental skills, professional ethics, client awareness, and case experience all must be learned. But you cannot use the previous generation's path to success as the sole determinant of your direction for the next decade.
You must reason backward from the future to today: which industries will continue to grow, which clients will have long-term legal needs, which capabilities will be replaced by tools, and which capabilities will become more valuable because of tools.
What is truly worth investing in may not be the hottest case sources of the moment. An industry with long-term growth that also demands sustained professional accumulation is more deserving of a young lawyer's dedicated commitment over several years. Slowly mastering the business, the technology, the regulation, the client's language, and the global market may not generate returns the fastest, but it is more like a long-term career asset.
Entrepreneurship Is Always Day One
What Mankun Law Firm aims to do is to face the future with the open mindset of a startup's first day, while doing the present work well.
We focus on Web3, AI, and fintech not because these terms are new, but because clients in these industries are encountering real problems in the global market. Where to incorporate, how to ensure business compliance, how to handle cross-border data transfer, how funds should flow, how to apply for licenses, how to explain products, how to control risks, and how to handle disputes when they arise—none of these can be fully addressed by single-point, offline, one-off legal services.
It requires a legal services organization that stays continuously online, collaborates rapidly, understands the business, can mobilize resources across regions, can build trust through content, and can accumulate experience through systems.
So the law firm of the future, as I understand it, is not about slapping on an AI label, buying a pile of software, or moving a traditional law firm's filing cabinets to the cloud. It is more like a new kind of organizational capability: an online gateway for client trust, an internal collaboration system for the team, AI-powered support and companionship in client service, the amplification of individual lawyers' abilities, and the long-term accumulation of the firm's knowledge.
This is also why we welcome forward-looking legal professionals to join Mankun.
Mankun now has branches in Shanghai, Shenzhen, Hong Kong, and Silicon Valley. We especially welcome lawyers in these cities who aspire to practice new-economy law for the global market to continue exploring, together with us, the professional possibilities for Chinese law firms and lawyers.
If you are willing to work in Web3, AI, and fintech, willing to serve entrepreneurs around the world, committed to a specialized and professional development path, and curious about what possibilities the legal industry offers beyond traditional routes, then we can work hard together.
Our generation of legal professionals is both fortunate and burdened. We are fortunate that changes in technology, industry, and global markets have brought many new demands for legal services to our doorstep; we are burdened because the old paths remain while the new paths have yet to fully take shape, and many things require us to step in and immerse ourselves.
This path will not be easy, but it may be what our generation of legal professionals truly ought to do.
