The managing partner and tech CEO speak to why they made the announcement and what it is meant to do
Generative AI is on track to become as basic to law practice as the smartphone, according to McMillan LLP CEO and managing partner Bruce Chapple. To that end, his firm recently announced the firmwide adoption of Legora, an AI platform built specifically for legal professionals.
“We want all of our team, all of our business professionals, everyone to be learning this new, state-of-the-art technology, because it’s going to be important for our competitiveness going forward,” he says.
For Legora CEO and co-founder Max Junestrand, the mandate is to move well beyond the browser-based chatbots that everyone, including lawyers, has been experimenting with over the last few years. He points to Legora’s tabular review feature as an example, “used for due diligence and discovery exercises, where you need to review hundreds or thousands of documents at the same time, and a chat just doesn’t cut it there.” The platform is built to sit inside the tools lawyers already use, including Microsoft Word, which he calls “crucial, to review and to draft, and to meet the lawyers where they are.”
He also describes the law firm’s AI stack as a hierarchy of responsibility. “If you think about the pyramid: at the bottom you have the foundational model; the second layer is the legal interpretation of that, which is Legora; and at the top of that, you have the McMillan configuration of these systems,” he says, arguing that the real value sits where firm-specific judgment rides on top of generic models.
Chapple says the technology only matters if lawyers actually use it, which is why McMillan’s selection process became a live test rather than a static procurement exercise. The firm piloted several tools, put Legora head-to-head with a rival and opened access to 100 users. Over the trial, usage did not plateau; it accelerated. “Legora came back and told us, ‘Your users’ usage is going up during this period.’ So it wasn’t someone trying it once and saying it does or doesn’t work. They were actually, in a very short period, becoming dependent on it, finding new ways to use it, and getting excited about it,” Chapple says.
That response gave him the confidence to push for a firmwide rollout rather than a narrow pilot. Scaling that usage, he adds, means treating AI as a process issue rather than just a technology purchase. Inside Legora, Junestrand says, “We think about use case development in two parts. One is getting everybody AI-literate and comfortable with the tools. The second is a top-down process reengineering in the biggest practice areas where there’s the most work. We rebuild the process from the ground up with an AI-first mindset,” he says.
Those redesigned processes are already emerging in high-volume work. Chapple says litigation and regulatory teams dealing with “a huge volume of documents to look at” are using AI “to look at those documents and pull out certain pieces of information across thousands or tens of thousands of documents,” while deal lawyers apply the same approach to contract-heavy due diligence, where speed and scale have always been pain points.
McMillan is also finding use cases in drafting. Chapple says lawyers are using the system to generate initial drafts of documents, such as affidavits, that they can then refine. “Often, what the system produces needs refinement, but that is actually a good thing. I continue to value the human oversight,” he says. “Even if only humans were involved in that process, there could still be mistakes. We have to observe our professional responsibilities.”
For Junestrand, the aim is not simply to save time on low-level work. AI can deepen the work while also changing how lawyers think. “It’s also important to recognize that in some cases, AI is not a time saver,” he says. “It can actually make it take longer to complete a task – but the quality goes up.” If a litigator uses the system as a thought partner, “bouncing ideas with AI might make you spend more time, but you may come up with two, three, or four new strategies you otherwise wouldn’t have.”
None of this matters if the product cannot clear the security bar. Junestrand says Legora made that call early and paid for it, “spending half of our initial angel funding getting all the certifications and building the right architecture from the very beginning to serve enterprise clients,” he says. That investment still underpins the roadmap. “We have four lead words in terms of how we prioritize product at Legora: secure is number one, transparent is number two, fast is number three, and cheap is number four.”
He is just as explicit about what happens to client data inside the platform. He stresses that no data is being used to train any underlying AI model, that nothing is visible to “any human reviewer or system with any of the underlying model providers or hyperscalers,” and that “none of the data is available to any other client.”
Chapple says skepticism among lawyers about AI is diminishing significantly as those safeguards and concrete use cases become clearer. “Even the most skeptical are now saying, ‘This looks pretty cool, some of the things that can happen here.’”
On Junestrand’s account, the collaboration between his company and its law firm customers is central to that shift. “We’ve been very clear from the start that to really change an industry, you need to do it together. We can come in as the technologists and the software engineers, but we need to partner with the best lawyers,” he says.
He also expects the tools themselves to move quickly up the value chain. “You can expect two things. First, the complexity of tasks that AI can complete end-to-end will continue to increase. NDAs were years ago. Now we’re talking about reviewing a data room. At some point, we’ll be talking about making a first draft of a [sale and purchase agreement], and it will be really good,” he says.
Chapple is watching a different set of indicators. For him, the question is not how impressive the technology looks in a demo but whether it is changing daily behaviour. In practice, that comes down to adoption, usage, greater client involvement, and innovative ways to use the tools. “We are literally, every day, figuring out new uses and new ways to work with team members and with clients. We are looking for solutions that are faster, more effective, and that draw on a lot of data. To do that, we need to include more people in that collaborative effort.”
This article is based on an episode of CL Talk, which can also be found here:
The episode can also be found on our CL Talk podcast homepage, which includes links to follow CL Talk on all the major podcast providers.
