Billing, disclosure, and the future of the small firm · AI for Lawyers

A senior advocate in a Babarmahal firm finishes a piece of work on a Tuesday evening in two hours that, eighteen months ago, would have taken her eight. She has done it well; the verification was thorough; the client will receive something at least as good as what they would have received last year, and possibly better, because she had time to think rather than time to type. She now stares at a half-completed time sheet. Eight hours is what the rate card and the long habit of the profession expect. Two hours is what the work took. She has to write a number in a box, and whichever number she writes is a quiet statement about what kind of firm she thinks she is running, and what her practice is actually selling. This final section is about that decision. It does not pretend to give one right answer. It does insist that there are wrong ones, and that the moment of choice is not somewhere over the horizon. It is on the desk in front of her, tonight.

The honest framing nobody wants to start with

Hourly billing in legal practice rests on a simple promise. The client pays for time. The advocate sells time. The rate per hour is the negotiated meeting point, and the time sheet is the bridge between them. The promise made sense in 1985, and it made sense in 2015, and it does not, quite, make sense in 2026 — because the unit being sold and the unit being bought have come apart. A drafting task that took a competent senior eight hours in 2015 takes a competent senior with a well-set-up AI workflow under two. The output is the same. The legal judgement that went into it is the same. The verification — done properly — is the same. The time, which is what the bill is denominated in, is a quarter.

If the senior writes eight hours on the time sheet anyway, she is — let us be plain — telling a quiet lie. The lie is not enormous on any individual brief. It compounds. Eventually a junior who joined the firm in 2024 will notice that partners are billing eight hours for tasks she finishes in ninety minutes with the same tools, and she will have a question the partners will not have a good answer for. Eventually a client whose son works at a tech company in Bangalore will notice the same pattern from the other direction. Eventually a Bar Council inquiry, asking about overcharging in a fee dispute, will look at the file and see a brief that bears the unmistakable structural fingerprints of AI assistance, billed at eight hours of human time. The current way of charging cannot survive widespread AI competence at the associate level. The firms that pretend otherwise will be the ones whose best juniors leave first and whose best clients leave second.

Three honest paths forward

There are essentially three responses that hold together under scrutiny. None of them is comfortable; that is part of why so few firms have moved yet.

One: bill actual hours, accept the revenue cut. The cleanest path. The senior writes 1.8 hours on the time sheet because that is what the work took. The bill to the client reflects 1.8 hours at the agreed rate. The firm absorbs the revenue compression on tasks AI accelerates, and tries to make up the gap by taking on more matters, by raising rates for the human-judgement portion of work, or by accepting a smaller but cleaner top line. The moral advantage is total — there is no conversation that begins with “actually, about that invoice.” The cost is real revenue, and any partner who tells you otherwise has not done the arithmetic. The firms that go this route will look smaller for a few years, then will look, by 2030, like the firms whose word means what it appears to mean.

Two: move to value-based or fixed fees agreed up front. The middle path, and probably the right one for most commercial engagements. Instead of billing time, the firm agrees a fee for an outcome — a fixed price for a particular kind of agreement drafted, a monthly retainer for a defined scope of advisory work, a staged fee for a particular litigation matter through specified milestones. The fee reflects what the work is worth to the client, not what it costs the firm in hours. AI productivity gains then accrue to the firm as margin, which is acceptable because the client got the agreed outcome at the agreed price. This requires real conversations at the start of an engagement; it requires the firm to know what its work is worth; it requires the courage to walk away from clients who insist on hourly billing as a cost-control mechanism. The reward is a practice that is not undone by every productivity improvement.

Three: carve out an explicit AI line item. The hybrid path. Time is billed at a reduced rate where AI did the heavy lifting, and the engagement letter shows a separate “technology and platform” charge that recovers the firm’s investment in enterprise AI subscriptions, training, and the secure machine in the back room running a local model for sensitive work. The client sees what they are paying for. The firm does not pretend the AI is free, and does not pretend the AI is irrelevant. This is the easiest path to explain to a long-standing client suspicious of any change to the billing model, and it is the easiest path to abuse, if the technology charge starts quietly outgrowing the value it represents. Used honestly, it is a real option. Used cynically, it is the old eight-hour bill with new packaging.

Any of the three is defensible. The fourth path — billing eight hours for two hours of work, telling nobody, hoping the question never comes up — is the one that ends, in a fee dispute or a junior’s resignation letter or a client’s quiet decision to move firms, badly.

The disclosure question

There is a related question that every firm will face from clients, and increasingly from courts, over the next two years. Did you use AI on my matter, and if so, where? The answer cannot be the marketing-flavoured non-answer of “we use a range of modern tools.” It has to be a real position.

The position that holds up is this: disclosure scales with risk. AI used for an internal first-draft note that a senior then rewrote? No client disclosure needed; the work product is the senior’s, and the AI is no more remarkable than a word processor. AI used to summarise a long bundle that informed advice given orally? Probably no formal disclosure, but the engagement file should record it in a single line. AI used to draft the actual text of advice that the client will rely on in a transaction or a filing? Yes — the file should record it, the partner should know it, and if the client asks, the firm should say so plainly. AI used in any way that shapes a pleading, a clause in an agreement that goes into force, or the contents of a court filing? Always recorded, with the verification step also recorded, because that is the workpaper trail a Bar Council inquiry, a successor counsel, or a court will look for.

Some jurisdictions are now formalising this. Judges in parts of India, the UK, and the US increasingly require counsel to certify whether AI was used to prepare a filing and whether citations have been verified. Nepal has not formalised the requirement yet, but the direction of travel is unmistakable. The firms that build disclosure into the file note today are the ones that will not have to scramble in 2027 when a rule lands.

The mistake to avoid is the opposite extreme — disclosing every keystroke of AI use as if to inoculate against future complaint. That tells the client they are paying for a process they no longer understand, and invites exactly the conversation about pricing the firm is trying not to have. Real disclosure is calibrated. It says, here are the places AI shaped the work, here is how we verified it, here is why we are confident in the result.

The market is already moving

It is tempting, in Nepali practice, to assume the AI conversation is several years away — that clients here will not notice, that competitors here will not change, that the existing way of working has another decade in it. This is wrong, and demonstrably so in 2026. The mid-tier firms in Kathmandu are already using enterprise tools. The juniors entering the profession from Tribhuvan University, Kathmandu School of Law, and the foreign LL.M. circuits are fluent in chat interfaces and arrive expecting their seniors to be too. The clients — particularly the export-oriented manufacturers, the SaaS founders, the second-generation business owners who studied abroad — already know what AI is and have a working sense of what it can and cannot do.

What they do not have yet, in most cases, is a coherent story from their lawyer about it. The firm that gets the story right first — “we use AI in these specific places, here is our verification protocol, here is what we will not use it for and why, here is how our fee reflects judgement and accountability rather than typing speed” — will look like the serious profession. The firm that has no story will look like the firm that did not notice. In any market with even moderate competition, that gap closes work within three to five years.

The Nepali clients who matter most for a firm’s next decade — the growing commercial businesses, the cross-border deal flow, the institutional clients with their own legal teams — will, in 2027 and 2028, be choosing counsel with this question in mind. Not “do you use AI” — they will assume you do. But “do you have a position on it that you can defend in a meeting.” A firm without a defensible position will not be on the shortlist.

The future of the small firm

It is worth ending the course on a more direct note than the practical sections so far.

The mechanical parts of legal practice — the first draft of a routine agreement, the summary of a long bundle, the digest of a regulatory notice, the initial research note on a familiar issue, the formatting of a standard letter, the comparison of two versions of a clause — have, for almost the entire history of the profession in Nepal, been most of the working day. They were what juniors did to learn the craft, and what mid-level associates did because there was simply nothing else to do with the hours. AI changes that. Not all of it, not overnight, but in the direction of less and less of the day being mechanical.

What is left — what AI cannot do, and probably will not do for a long time — is what the profession was always claiming to be about anyway. The judgement to know when a clause is aggressive enough to warrant a real conversation with the client before they sign. The instinct to read a bench and adapt an argument mid-hearing. The relationship that means a businessman calls you before he signs the joint-venture, not after. The trust that means a counterparty’s solicitor accepts your representation of a fact in negotiation because of the name at the bottom of the letter. The judgement to know which arguments to make, which battles to concede, which clients to advise to settle. These are not mechanical things. They are the things human beings — with reputations, with bar licences, with a stake in being the kind of advocate other advocates describe with respect — are uniquely placed to provide.

That is what you charge for. Not the typing. Not the search. Not the first draft. The judgement, the relationship, the trust, the accountability before a bench and a Bar Council and a client. AI has, in a quiet way, done the Nepali legal profession a real favour. It has stripped away the mechanical work that was distracting from the actual thing — the work the profession was always supposed to be doing, and was, in busier weeks, often too tired to do well. The advocates who notice — who reorganise their pricing and their practice and their training of juniors around what AI cannot do, rather than competing on speed at what it can — will not just survive the next decade. They will look, in 2035, like the people who first understood what the practice of law in Nepal was always supposed to be.

That is the bet this course has been making, from the first section to this one. It is the bet worth making.

What comes next

This was the final section of AI for Lawyers. What remains is the course exam — ten questions drawn from across the full course, covering everything from where AI fits in a Nepali small practice, through research and citation, contracts, client work, pleadings and submissions, and the privilege, verification, and billing material of this final chapter. A score of 80% is required to pass. The exam is here: /courses/ai-for-lawyers/exam. Take it when you have a clear half hour and the course material fresh in mind.