Two years into my career as a disputes lawyer at a leading City firm, I asked myself a question that changed everything: what makes a successful partner in ten years' time?
The answer is not more billable hours, a narrower specialism, or a bigger contacts book. The successful partners of the future will be those who have learned to leverage technology so completely that it has fundamentally altered who they are as lawyers - not just the speed at which they work, but the quality of their thinking, the breadth of what they can take on, and the value they deliver to clients. I realised that the lawyers who would thrive were not the ones dabbling in AI. The lawyers who would thrive were rebuilding their entire practice around it.
That realisation set me on a path from practising law to helping build the technology I'd become obsessed with. But this piece isn't about me. It's about you - and the mental model shift that will separate the lawyers who thrive in 2030 from those who get left behind.
Use cases are just the beginning
As a Legal Engineer, a large part of my role involves running use case discovery sessions with the lawyers who use the Legora platform. The objective is straightforward: diagnose friction points in a lawyer's practice, identify which ones AI can address, and build solutions accordingly. These sessions are truly valuable. They give lawyers who are new to AI a concrete starting point, a reason to engage, and real momentum.
But here's what I've observed: for many lawyers, use cases become the ceiling rather than the floor. They attend a session, compile a list of four or five things the tool is good for, and then reach for AI only when one of those scenarios arises. Document review? Sure. Issues list from an SPA? Absolutely. But everything else continues exactly as before - the same processes, the same bottlenecks, the same way of practising law that hasn't materially changed in decades. Use cases are the perfect on-ramp. The problem is when lawyers treat them as the destination.
This is the classic faster horse problem. When Henry Ford asked people what they wanted, they said a faster horse. They couldn't conceive of the car because they were trapped in the mental model of what already existed. And that is exactly what is happening in law right now. We are retrofitting old ways of working onto a technology that demands an entirely new operating model. We're trying to bolt a jet engine onto a horse-drawn carriage and wondering why it doesn't fly.
My time in practice
When my previous firm rolled out Legora, I threw myself into it - not tentatively, not with a list of approved use cases, but completely. My mindset was simple: whenever a new piece of work came in, I would assume that the platform could help. I would open it before I even knew how I was going to use it. I wasn't deterred when it took time to figure out the right approach. I treated every task as an opportunity to push the boundaries of what was possible.
Within months, Legora was the first thing I opened in the morning and the last thing I closed at night. It wasn't a tool I reached for occasionally. It was my legal operating system: the place where all of my work lived, where every document, every matter, every piece of context was connected and accessible. Agentic systems, which were not available to me in practice, elevate this dramatically, reducing the onus on lawyers to know and select the right tool, and shifting their role to that of direction, supervision, critical analysis, and as the ultimate arbiter of what work leaves the building.
People in practice would often come up to me and ask: what are your prompts? Can you share your use cases? And of course I would, but if we're having a prompt-sharing exercise, we are in danger of missing the point. Sharing prompts or use cases implies there's a finite set of things AI can do and a larger set of things it can't. That's the wrong mental model. The question shouldn't be 'which of my tasks can AI help with?' It should be 'which of my tasks can AI not help with?' - and you'll find that list is vanishingly small.
Old world vs. new world
Consider a typical piece of legal work today. Instructions arrive by email. You save the documents to the file. You open a blank document and start drafting from scratch, maybe from memory, maybe pulling up a precedent if you can find one. You refine your drafting, perhaps checking it against a playbook. You send a draft to a colleague, wait a day for their marked-up response, incorporate their changes, and draft an email to the client with the document attached. Each task is handled sequentially, one instrument played at a time, one note following the last.
It works. But look at what it costs. The output is inconsistent: you're drafting in isolation from the firm's best work, so quality varies from lawyer to lawyer, matter to matter. It is slow because every step waits on the last, and collaboration happens in fits and starts. And it can be crushing from a personal perspective. Your day is consumed by the mechanical grind of document processing, leaving little room for the strategic thinking, the creative problem-solving, and the client relationships that drew you to the law in the first place. Your to-do list is relentless. Your mental bandwidth is exhausted before you get to the work that actually matters.
Now picture the alternative. Instructions arrive and you save them to your collaborative AI workspace - a single AI-native environment that already contains every document and piece of context on that matter, connected to your firm's full body of institutional knowledge. The platform generates a holding response to the client and automatically updates your agentic task list with the instructions. You orient yourself, and then you brief the Agent as you would brief a capable team member: the outcome, the commercial context, the constraints, the sources it must use and the form the final deliverable should take. It proposes a plan before it acts. You can challenge the scope, redirect a workstream or add an approval point, and then let it execute.
From there, the Agent can draw on the different parts of the operating system, along with the full history of the file, prior correspondence and your firm's gold standard precedents, without waiting for you to move between them. For a complex task, it can divide the work into parallel lines of inquiry and reconcile them into one coherent output. While that is happening, you are thinking strategically about the commercial implications, or directing other workstreams. You and your colleagues are called in at the points where judgement is required, to verify the sources and reasoning, refine the draft, and approve what goes to the client.
Where the old world was sequential - one instrument at a time - the new one can be symphonic. Multiple workstreams move in parallel. Context flows between them. The quality of output is dramatically more consistent, because every draft is rooted in the best the firm has ever produced. The speed is transformative, because you are no longer the bottleneck in your own workflow. And the human cost inverts entirely: instead of wading through documents, you have the breathing space to think - about strategy, about commercial context, about the implications that your client is actually paying you to consider. You are no longer playing every instrument yourself. You are conducting the orchestra.
Unlocking institutional memory
There is another dimension to this shift that goes beyond any individual lawyer's productivity. Every law firm sits on a vast body of institutional memory - years of precedent documents, strategic advice and hard-won judgement calls - locked away in document management systems, buried in email chains, or scattered across shared drives in formats that make it almost impossible to find when you need it. Ask any lawyer whether they've ever spent an hour searching for an example of something they know the firm has done before, and they'll laugh at the understatement.
When a platform like Legora becomes your firm's operating system - when every matter, every document, every piece of work product flows through it - something remarkable happens. That institutional memory becomes searchable and available to every lawyer in the firm at the moment they need it. A junior associate drafting an indemnity clause at midnight can draw on the best example the firm has ever produced, informed by the strategic reasoning of the partner who negotiated it five years ago. The AI retrieves it, works out what's relevant, and helps the lawyer apply it to the matter at hand.
This is a compounding benefit. The more deeply a firm embeds AI into its practice, the richer and more accessible that institutional knowledge becomes. Every piece of work feeds the system. Every matter adds context. Over time, the platform can encode the expertise of the entire firm and make it collectively smarter. The gold-standard work product of your best partners becomes the baseline for everyone. That is a firm-wide transformation that goes far beyond a simple productivity boost.
The lawyer of 2030
Leaving the institutional benefits to one side, this is a profoundly human story. Every lawyer I speak to is wrestling with the same tension: they know AI is important, they've been told to adopt it, but nobody is actually talking about what this means for them - for their career, their identity, their future in the profession. Everyone is discussing the technology, the products, even the new fee structures. But nobody is addressing the question that actually keeps lawyers up at night: what do I need to become?
So let's address the elephant in the room. Will AI replace me? Will it commoditise my expertise and eat into my margins? These are legitimate fears, and I don't think the profession has been honest enough in confronting them.
Here is what I believe, grounded in my own experience on both sides of the aisle: AI will not steal your expertise. It will amplify it. The work that AI takes off your plate is the mechanical, repetitive work that no lawyer ever enjoyed doing and no client ever wanted to pay a premium for. What it gives back is something far more valuable - the time and mental clarity to exercise the judgement, the strategic thinking, and the creative problem-solving that are uniquely human. The work that made you want to be a lawyer in the first place.
I know lawyers who left the profession entirely - talented people, exhausted by the grind. Some of them then witnessed how deeply I was incorporating AI into my practice and told me they would never have left if they'd had these tools available. That tells you everything about what this technology does to the lived experience of practising law. It doesn't diminish the lawyer. It liberates them.
The lawyer of 2030 is not someone who has memorised a list of AI use cases. They are someone who has undergone a complete identity shift, from operator to conductor. Think back to that orchestral metaphor. In the old world, you were a musician playing every instrument yourself, one at a time, trying to produce something that resembled a coherent performance. The lawyer of 2030 stands at the podium. They have the full score in front of them: the client's commercial objectives, the strategic context, the institutional knowledge of their firm.
The Agent becomes the orchestration layer between their direction and the work being done. The lawyer defines the objective and the non-negotiables. The Agent turns that brief into a plan, deploys the right capabilities and parallel workstreams, keeps track of progress, and brings the outputs back together. The lawyer does not disappear from the process; they intervene where their judgement has the highest value: approving the approach, resolving ambiguity, challenging conclusions, checking authorities and making the final call. The output isn't just faster. It is richer, more layered and more coherent than anything a single pair of hands could produce alone.
This also changes the relationship between attention and execution. Today, work generally stops when the lawyer's attention moves elsewhere. In an agentic model, properly scoped execution can continue while the lawyer is in a client meeting, considering strategy or supervising another matter. Attention is still precious, but it is concentrated at decision points rather than consumed by every mechanical step. The defining skill becomes the quality of your direction: knowing what to delegate, how to scope, what evidence to demand, and when to intervene.
I understand why this is difficult. Lawyers are trained to work within the confines of rules and precedent. We are not, by nature, the most instinctive innovators. And for years, we've had technology solutions thrust upon us - piecemeal, disconnected tools that solved narrow problems and never quite delivered on their promise. It's no wonder the instinct is to approach AI with the same scepticism and the same narrow framing.
But this is not another point solution being bolted onto your existing workflow. This is a completely new operating system for legal work that fundamentally changes what a single lawyer is capable of producing. To treat it as anything less is to leave extraordinary potential on the table.
This is revolution, not evolution
AI will not replace lawyers. But lawyers who truly embrace AI - not as a collection of use cases but as the foundation of how they work - will outperform those who don't. And the gap will only widen. This technology is going to allow litigators to take on more cases. It will allow corporate lawyers to run more transactions, more quickly. It will open up a vast body of legal demand that has gone unserved because the economics of traditional practice couldn't support it. The legal pie will get much bigger - and the lawyers who have shifted their mindset will claim a larger share of it.
The opportunity is extraordinary, and it is open to anyone willing to seize it. It doesn't take much to be a lawyer who uses AI more effectively than most of your peers; the bar, right now, is remarkably low. But it requires you to stop retrofitting AI onto a handful of existing processes and start thinking about a strategic redesign of the way your work gets done.
There is no roadmap for the future of legal practice. Nobody has done this before. We are at the frontier, and I understand that is daunting. But if my experience has taught me anything, it's this: the lawyers who will define the next decade are not the ones waiting for permission or a pre-approved list of use cases. They are the ones who have picked up the baton, stepped onto the podium and started conducting.
