When AI shrinks time, a lawyer’s billable hour becomes harder to defend

For Singapore’s legal profession, artificial intelligence is not an issue that lurks in the distance.

The disruption is already here and the sector is trying to come to grips with the impact of generative AI.

In his 2026 Committee of Supply speech, Minister for Law Edwin Tong touched on topics that the legal profession faces: who bears responsibility when AI makes errors, how young lawyers are to develop judgment if basic tasks are automated, and how regulation should keep pace without crushing innovation.

Those are important questions. But there is another one that law firms may find even more uncomfortable, because it goes to the heart of legal practice and encompasses how the services are delivered and ultimately how clients are billed.

If AI sharply reduces the time needed to perform legal work, how much longer can firms justify billing by the hour?

For decades, the billable hour has been the organising principle of private practice. It is not merely a pricing method; it shapes staffing, promotion, training, profitability and even status. In many firms, hours are the scoreboard. They tell partners who is productive, tell associates who is safe, and tell clients what they are being charged for.

But AI is beginning to expose the central weakness in that system: It rewards time spent, not value delivered.

What is a client paying for?

A lawyer who takes six hours to review a document set can traditionally bill more than one who takes three. If AI reduces that same exercise to 30 minutes with human supervision, the old equation starts to look untenable. The client will ask a perfectly reasonable question: Am I paying for legal judgment, or for the inefficiency of an older business model?

This is where the debate becomes sharper than many in the profession may like. Lawyers often say clients pay for expertise, not for keystrokes. If that is true, then many firms should stop pretending that time is still the fairest proxy for value.

The billable hour once had a practical logic. Legal work was labour-intensive, research-heavy and difficult to standardise. Time offered an administratively simple way to price uncertainty. But AI is rapidly weakening that justification. When software can search authorities, summarise large records, produce first drafts and flag issues across jurisdictions at speeds no human team can match, billing strictly by elapsed time starts to look less like professional discipline and more like institutional inertia.

This tension is especially relevant in Singapore, where the legal sector is under pressure to be both globally competitive and locally trusted.

It is a pressing issue for firms such as RHTLaw Asia, a Singapore-based, multidisciplinary professional services platform serving clients across Asia. Its clients are likely to become less tolerant of time-heavy billing for work that technology can accelerate. Cross-border businesses do not merely want legal labour; they want speed, judgment, coordination and commercial clarity. That is why the real threat from AI is not that it will replace lawyers. It is that it will make some lawyers’ pricing models look outdated.

The profession should be honest about this. AI does not simply create efficiency gains. It creates a pricing crisis for firms that still depend too heavily on the arithmetic of hours multiplied by rates. It may not be apparent now but the more capable the technology becomes, the more exposed that model becomes. A firm can either pass some of the efficiency gain to clients, or attempt to preserve old economics while using new tools. The second option may be profitable in the short run. It is also likely to be the one that damages trust.

And trust is not a side issue in law. It is the business of law.

Hybrid pricing?

Some lawyers will argue that billable hours are still necessary because legal matters remain unpredictable. That is true, especially in complex disputes, regulatory investigations and transactions where the scope can evolve quickly. The answer is not that every matter should move overnight to fixed fees. The answer is that the profession can no longer defend the billable hour as its default moral centre.

The more credible future lies in hybrid pricing: fixed or capped fees for clearly bounded work, phased fees or milestone-based billing for transactions, retainers for ongoing advisory, success-linked components where appropriate, and premium pricing for high-stakes judgment as well as project work rather than low-level drafting.

Yet the path to that future is neither smooth nor uniform. Professional conduct rules prohibiting overcharging will come into play. Now that AI can research, review, and draft in a fraction of the time previously required, the “time necessary” to do the legal work has objectively changed. The issue gets clouded further with the profession adopting AI at uneven rates.

Early adopters face a paradox: their investment makes them faster, but hourly billing penalises them for it. Laggards may earn more per matter in the short term simply by being slower, but soon their clients will begin to ask not just how long a task took, but how long it should have taken. Firms that thrive will increasingly be those that price for insight, assurance and accountability, not for drafting time alone. Lawyers who deliver solutions, anticipate roadblocks and maintain strong client relationship can command higher fees while remaining fully compliant with rules against excessive billing.

There is also a harder truth. The billable hour has survived not only because clients tolerated it, but because it served internal law firm economics. It funded leverage, monetised junior labour, and gave firms a tidy way to measure performance. AI now threatens all three.

This is why the junior lawyer question matters so much. If AI handles first drafts, document review and much of the repetitive research once done by associates, then the profession cannot simply celebrate productivity and move on. It must confront how future lawyers will be trained, assessed and promoted. Minister Tong put the issue plainly: If AI performs the basic tasks and “gives the answers”, how does a young lawyer develop instinct and judgment? If clients will no longer be willing to pay junior lawyers for grunt work that AI can do, then the apprenticeship model of law must change too.

That means firms will need to invest more in client-facing responsibility and scenario-based training for junior lawyers. In other words, they may have to spend more on training even while billing fewer hours. Some will resist this. But resistance is not a strategy.

The legal profession often prides itself on being conservative, but there is a difference between prudence and denial. AI has already made its mark. The market will move next, and clients may prove less sentimental about the billable hour than lawyers are.

So the question is no longer whether AI will affect billing in law.

It already has.

The real question is whether firms will redesign their economics before clients, competitors and technology force the issue upon them and threaten their profitability and sustainability.

Because when AI shrinks time, the billable hour does not just become inefficient.

It becomes harder to defend.