“Patent work is one of the best places to evaluate the impact of AI. The workflows are precise, the deliverables are concrete, flat-fee structures are common, and the relationship between AI investment and outcomes is unusually visible.”
I have spent most of my professional career talking to patent practitioners about AI. For years, the conversation was about whether AI could be trusted, whether it was ready, and whether it would actually change how patent work gets done. I have watched the profession move from skepticism to curiosity to cautious adoption to – in 2026, for the first time – something that feels like acceptance. Questions that once provoked heated debate at conferences now feel almost trite. Nobody is really questioning whether AI has a place in patent practice anymore. The question that has replaced it is harder and more consequential:
How do we measure the actual value of AI?
It is the right question. But most of the profession is trying to answer it with a stopwatch when they should be asking what AI has made possible that was not possible before.
A Shifting Dialogue
In 2025, the patent profession adopted AI. Organizations bought tools in droves – they ran pilots, built workflows and got comfortable with the idea that AI was here to stay. We crossed the chasm. But 2026 is the year the profession has to figure out what all of that adoption is worth – and the people paying for it are starting to ask the same question.
Most conversations around AI’s return on investment (ROI) default to time saved. Hours shaved off a prior art search, a draft application produced in minutes instead of days, an office action response prepared almost instantly, a due diligence review compressed from a week to an afternoon. These gains are real and should not be discounted entirely. But they are also the smallest part of the story – and building your entire AI value narrative around efficiency is a dangerous place to be, especially when clients are listening.
I was invited to present on this topic at LegalWeek in New York City last month. Having attended for the first time, I was not sure how the message was going to be received. Nearly 1,000 professionals showed up, and they were not just listening, they were nodding along. It became clear this was not just my observation; it was the entire industry’s. Panel after panel, the conversation would begin with efficiency metrics and then, almost inevitably, drift toward something bigger and harder to quantify. A partner describing how AI changed which clients he could serve and how. A CFO introducing a framework that reframed the entire value question. A firm leader admitting that the way they had been pitching AI to clients was actually undermining their pricing power.
The Real ROI of AI for Patent Work
But what became increasingly clear was that the broader legal industry was already having this conversation in sophisticated ways, and patent practice was barely part of it. That felt like a missed opportunity because patent work is one of the best places to evaluate the impact of AI. The workflows are precise, the deliverables are concrete, flat-fee structures are common, and the relationship between AI investment and outcomes is unusually visible. There was almost nothing out there examining this seriously for our profession. This realization, and a lot of subsequent research and conversations, became the basis for a white paper I have just published: The Real ROI of AI in Patent Practice.
The white paper goes into considerable detail across the data, the frameworks, and the practical implications for patent practitioners. But there are two ideas in it that I think are worth sharing here.
The first is that the legal industry is sitting on an iceberg and most firms can only see the tip. BigHand’s 2026 Annual Law Firm Finance Report, one of the largest surveys of its kind, covering more than 800 senior finance and legal professionals across the general legal industry, tells a story that is frankly a little unsettling. The headline numbers look promising; the numbers underneath them do not. The traditional model is under structural pressure in ways that rate increases and billable hour targets cannot fix, and AI is accelerating that pressure, whether firms are paying attention or not. Patent practice is no exception.
The second is a framework that came up during my panel at LegalWeek, introduced by fellow panelist Madhav Srinivasan, CFO at FBT Gibbons, who presented a terrific yet simple categorization for measuring value. He sorts AI value into three buckets: Quantity, Quality, and Quantum. Most firms are measuring Quantity – time saved, tasks completed faster. A few are starting to measure Quality – better work product, fewer errors, higher caliber output. Almost nobody is measuring Quantum – the truly transformative impact, unlocking ways of working that were not possible before.
Continue the Conversation
The white paper explores what Quantum looks like for patent practitioners, why the firms that lead with efficiency are playing a commercially dangerous game, and why the true cost of getting AI implementation “right” goes far beyond the tools themselves. Having spent my career at the intersection of AI and patent practice, and having seen firsthand how the broader legal industry is in thinking about these questions, I felt uniquely positioned to tailor them for patent practitioners.
If these questions are on your mind, I would encourage you to read the full report. We also explored many of these themes in a recent IPWatchdog webinar, which you can watch here. I would love to have your comments below and continue the conversation.
