The Three-Hour Riddle: Who Gets the Time AI Saves a Law Firm?

Here’s a small riddle for anyone who runs a law firm. A first draft that used to take an associate four hours now takes one. Where do the other three go?

Most of the talk about AI in law skips straight to the answer: back into the machine, as more matters carried by the same people. Fair enough. It’s a perfectly respectable business decision. But it is a decision and I think the firms that give some of those hours back to their people, as thinking time, client time and actual evenings, are placing the smarter long-term bet.

Space, not surplus

The urge to refill the hours is understandable. Nature abhors a vacuum and so does a billing target. But imagine treating that time as breathing room instead. The associate gets to think a matter through rather than draft with one eye on the clock. The senior attorney gets an unhurried half hour with a client who needed it. Everybody gets home while it’s still light out, at least some of the time.

None of this is charity. It’s a wager that rested people think more clearly, make fewer mistakes and stick around longer, and that the pay-off arrives over years rather than by the next invoice run.

The hours nobody will miss

Let’s be honest about which hours AI is actually taking. It’s assembling first drafts from templates, summarising transcripts, combing disclosure bundles for the three pages that matter out of three hundred, lining dates up against a limitation period. In my experience, this is the work behind most late nights in practice. It is not, I’d wager, what anybody pictured the first time they imagined themselves in a gown.

Lifting it off the desk doesn’t make the lawyer redundant. Somebody still has to check the draft and confirm the summary caught what counts. The hours simply move, away from producing the rough first pass and towards the judgement that rough pass always needed.

Mind the ladder

Here’s where it gets delicate. Generations of lawyers learned the craft precisely by doing this unglamorous work, year after year. Take it away carelessly and you risk promoting juniors into client-facing advice before they’ve built the instincts the grind used to give them. It’s a bit like handing someone the car keys because the car can park itself.

The firms that get this right will still make associates learn the fundamentals properly, by hand, before they’re allowed to lean on AI. Once that foundation has set, though, those associates can move on to meatier work sooner, because the routine tasks are quicker to produce and easier to check.

Handled badly, the same change becomes a shortcut around the training instead of a tool that comes after it. Which version a firm ends up with depends on how carefully it designs the transition. The software won’t make that call for you.

The quiet dividend

Many firms, in my experience, treat losing junior lawyers as something close to a law of nature. If AI clears away the repetitive work that wears people down, the firms that share that windfall with their people, rather than swallowing it whole, should find recruiting and keeping good lawyers noticeably easier.

And the benefit snowballs. Lawyers who stay build up institutional memory, deeper client relationships and the kind of judgement only time can grow. The firm saves the cost of constant replacement and gains something clients notice straight away: a familiar voice on the other end of the phone.

Back to the riddle

So, those three hours. The software that freed them up has no opinion on what happens next. It will fill them with billable work or hand them back with a bow on top, equally cheerfully. Someone in the firm decides and most firms will decide by never asking the question. Squeeze the hours for output and you’ll get more output for a while, probably alongside a steady trickle of resignation letters. Give some of them back and you’re betting on something no competitor can simply buy: people who want to stay.

Javed E. Daniel (LL.B) (LEC)