The billable hour is evolving. Here’s what the patent industry is getting wrong about it

Most patent attorneys didn’t go looking for a billing crisis, but the conversation found them anyway. Clients armed with AI tools are pushing back on fees, industry forecasters are declaring the hourly model obsolete, and a low-level anxiety has taken hold: if AI cuts the time it takes to do the work, does it also cut the revenue?

It’s a reasonable fear, and for some firms it’s taken a specific shape: if AI compresses the time it takes to do the work, how do you justify the rate to a client who knows it? That tension is real, but the conclusion most firms are drawing from it—that the billable model itself is the problem—is where the reasoning goes wrong.

90% of legal professionals now believe that generative AI has already significantly altered billing practices or will do so within two years.

For patent attorneys whose revenue depends on hours billed, that statistic carries real weight. But the dominant narrative that AI time savings translate directly into billable hours lost rests on a misreading of the evidence.

AI time savings and patent billing: what the data shows

The most cited evidence in the billable hour debate is also the most misinterpreted. When researchers report that AI saves attorneys hours, the assumption is that those hours were billable and their disappearance represents lost revenue.

Thomson Reuters’ 2024 Future of Professionals Report surveyed more than 2,000 legal professionals globally and found that AI has the potential to save attorneys up to 12 hours per week by 2029.

Thomson Reuters then ran the math that most commentators skip: for a US attorney, those savings translate to nearly $100,000 in additional billable capacity annually. The story becomes no longer one of lost revenue, but about recovered capacity for work that actually demands an attorney’s judgment.

A separate Everlaw survey of nearly 300 legal professionals found that lawyers using generative AI report saving up to 260 hours per year. For firms paying attention, 260 hours a year is capacity waiting to be deployed.

Why time saved doesn’t mean billable hours lost

The billable hour panic rests on the largely untested assumption that the hours AI saves are hours currently being billed. For most patent practices, that may not be where the time is going.

Among legal professionals who have widely adopted AI, 69% report a positive revenue impact, according to the 2025 Clio Legal Trends Report.

Two factors explain why:

1. Where AI actually lands in a practice

A substantial portion of attorney time is consumed by work that never makes it onto a timesheet: coordination overhead, document management, repetitive research tasks that don’t warrant a line item. AI addresses those tasks first. The result is more attorney bandwidth for the work clients actually pay for.

2. Throughput

A patent attorney who can draft an office action response in two hours instead of six has not lost four hours of revenue. They have gained capacity for the next matter.

Fixed-fee billing in patent practice: where it works and where it doesn’t

Fixed fees for routine patent prosecution are becoming an increasingly common client expectation, and AI makes it more viable to deliver on that promise without giving away margin. Firms offering alternative fee arrangements are already reporting 30-40% improvement in realization rates, according to LeanLaw’s 2025 IP billing analysis.

For firms still operating on hourly billing but weighing a future shift, AI adoption now is also the smartest preparation for that transition. Firms that are adopting AI into their workflows while billing hourly arrive at fixed-fee pricing with their margins already protected because the efficiency gains are already baked in. The move becomes viable precisely because the hard work of compressing task time has already been done.

However, complex prosecution work does not become predictable because a client prefers a fixed price. As IPWatchdog reported in May 2026, fixed fees depend on specific conditions holding: a manageable disclosure, a reasonable review cycle, and targeted client feedback.

When those break down (and in patent prosecution, they frequently do), the pricing model must flex. Fixed fees and hourly billing are not competing models. They are tools, and the firms performing best are choosing deliberately where each applies.

The client conversation patent firms need to have about AI

The real billing risk for patent firms is not the model itself. It is what happens when AI-driven efficiency goes unaddressed in client relationships.

Bloomberg Law’s Legal Exchange identified the dynamic clearly in early 2026: when firms don’t proactively reframe what AI efficiency means, clients don’t sit back and appreciate the improved output; they ask for discounts.

Without a deliberate conversation about value, efficiency gains get absorbed as price concessions rather than recognized as practice improvements.

The attorneys holding their rates are the ones who can articulate what the time savings produced: a more thorough application, a faster response cycle, bandwidth for strategic advisory work that previously got squeezed out by drafting.

How the patent billing model is actually evolving

Patent practice is sorting itself into two tracks.

Routine, scope-defined work is moving toward fixed fees, while complex, judgment-intensive prosecution—where an experienced attorney’s technical read is the actual product—continues to command hourly rates.

AI makes it possible to deliver both tracks at a higher standard. For firms that understand that, the billing conversation with clients looks very different.