How McLeish Orlando uses AI in personal injury files — and why smaller firms can’t ignore it
If you’d asked Dale Orlando even two years ago whether AI would soon revolutionize his practice, he would have said no, there’s no sign of that. Fast forward to today, he’s consistently impressed with that the technology can do and how it’s improved — and he can’t imagine how he’d operate without it.
“Now, it’s almost a necessity to practice efficiently and serve your clients,” says Orlando, principal partner at McLeish Orlando LLP. “The advent of AI models is not only exciting; it's no overstatement to call it transformative.”
When Orlando, along with John McLeish, started the firm in 1999, they had a conversation around technology. They both agreed that one of the foundational pillars would be to stay on the cutting edge. At that time, not everybody had an internet connection — there were a few computers at the firm that had access — and email was relatively new. But for those paying attention, the trajectory was clear.
“At the time we thought, wow, this tech revolution that's happening is really going to change things — and you can see how much has changed in the last 27 years,” Orlando says. “We pride ourselves on making it a priority from Day 1 to stay on top of it.”
In what he describes as the “pre‑AI era,” Orlando says the firm easily kept pace with innovation. McLeish Orlando moved early into what was the big step at the time, digital document management. They started with a system called Primafact more than 20 years ago before migrating to Filevine, a cloud-based platform that centralizes calendaring, docketing, document management, and task assignments. Instead of staff dropping by his office to ask questions or logging every conversation manually, internal communications now run through Filevine’s notes system, automatically attaching those exchanges to the relevant file.
Building on that foundation, McLeish Orlando took a more formal step into AI about a year and a half ago by striking an internal AI committee. The group’s mandate was to sort through the growing number of tools and “AI opportunities” landing in their inboxes and identify what might actually work in practice.
At first, Orlando recalls, the technology wasn’t quite there. But within several months, the tools matured to the point where it made sense to adopt them firm-wide and invest in their cost. They brought on Filevine’s AI model, Lois, and the efficiency gains since then have been incredible, improving noticeably even over the short time they’ve been using it.
Because it sits inside the existing case‑management system, Lois can see everything in a file — medical records, expert reports, pleadings, internal notes — without any extra uploading or data wrangling. Users can “chat with” a file and ask for a detailed chronology, a summary of an expert report, or even cross-checking of a lengthy neuropsychological assessment against the file material to identify areas of weakness, inconsistencies, or misstatements of evidence.
Another recent example was a case involving a crumbling skull versus thin skull argument, meaning that the client had significant pre-accident health issues and the question becomes whether but for the accident the person would have been fine or if those susceptibilities and vulnerabilities meant decline was inevitable. This issue comes up in a lot of cases, and a clerk would have had to take ten years’ worth of pre-accident health records ranging from hundreds to thousands of pages and go through everything to highlight relevant entries. That would take an average of half a day; Lois does it in five minutes.
“It really has changed how we practice it in a large way,” Orlando says, adding that the impact isn’t limited to back‑office work.
Before a call with a client or a defence lawyer, instead of lengthy briefings from junior lawyers, Orlando can ask Lois for an at‑a‑glance update on case status. In seconds, the AI pulls up the latest: whether a client has returned to work, what functional limitations have been documented in the last six months, and whether any treating expert has provided a prognosis.
Lois also proved its value during a recent pre‑trial on a long‑term disability case with tens of thousands of pages on file. Although Orlando is familiar with the firm’s cases and always prepares by thoroughly reviewing the file at issue, the judge in this instance asked highly detailed questions — including the evolution of the plaintiff’s medication use — that would normally require an hour of digging. By pausing briefly to consult Lois, he was able to retrieve precise answers in about 10 seconds each time, leaving the judge with the impression he was extraordinarily prepared.
“I didn’t have the heart to say, it’s not me, Your Honour, it’s Lois,” Orlando quips.
Everyone’s heard the warning stories about lawyers showing up to court with an AI-sourced case reference in their factum that doesn’t exist. Perhaps more concerning, technology companies don’t seem to fully understand why the technology hallucinates in the first place. AI, especially open-web tools, can misinterpret facts — something that humans must play backstop against.
For McLeish Orlando, Lois is a closed loop which means there’s no hallucination risk: it can’t look outside the file. Also, any answers it gives include footnotes that will take the lawyer directly to document it used to come to its conclusion.
“Potential subtle factual inaccuracies still require diligence on the part of the lawyer to cross-reference and double-check, particularly if something seems off,” Orlando says, adding that the other major concern with AI use is confidentiality, data governance, and storage.
Although there are always risks, these can be managed. For example, the firm is currently working with FileVine to have all data stored in Canada because of client concerns over some being held in the US. That said, the company codes everything so that the data is encrypted and unreadable on the small chance it’s accessed by bad actors.
A more pressing issue might be the possible eroding of skills due to over-reliance on AI. While the clerks and lawyers are encouraged to use Lois, Orlando says they are specifically asked if the results were verified; did they read the reports and cross reference?
“If they’re only relying on AI, you only know what it tells you — there needs to be an independent review of the facts,” Orlando explains. “For example, maybe there are some innovative ways to make inroads with respect to a cross-examination of a particular expert, that AI just can’t identify.”
That said, the firm is far less concerned with how a simple chronology is created, versus when a senior lawyer requests a critique of a defence report or position.
“That's a different ballgame,” Orlando says of the latter. “I expect the lawyers and clerks to use their own independent judgment, supported and supplemented by Lois, but they must have a critical eye.”
While the firm doesn’t expect to cut headcount because of AI, they will be less likely to bring on new support. People should be able to handle larger caseloads without compromising the quality of their work, with more emphasis placed on the human-facing aspects of the job, such as client conversations and witness interviews.
At this stage, with lived experience as to its capabilities, “it would be naïve to say AI is simply a productivity tool driving efficiency,” Orlando adds.
For Orlando, one of the most compelling aspects of AI is its ability to narrow the gap between large, well‑resourced firms and smaller practices. At a bigger shop like McLeish Orlando, there are junior lawyers and law clerks to support file work and preparation, which helps senior counsel stay thoroughly briefed. By contrast, a sole practitioner or two‑person operation may find it far more difficult to stay apprised on every detail in every case. AI, he says, can help bridge that divide.
“I literally use it all day, every day — I'm constantly interacting with Lois,” Orlando says. “It’s incumbent on lawyers to be up to speed on cases, but keep in mind many are three, four, even five years old. There are literally tens of thousands of documents. It’s impossible to know every aspect and that’s where AI makes a considerable difference.”
From a cost perspective, Orlando doesn’t see it as prohibitive, and believes tools like Lois can either help smaller practices manage workloads in a way that preserves some work‑life balance or make them substantially better prepared by assisting with analysis and issue‑spotting that would otherwise be out of reach.
“People that aren't starting to go down this road are losing out on a massive opportunity to improve the efficiency of their practice,” Orlando says. “While it might take some longer than others, at some point soon, we’ll see 100 per cent adoption. Lawyers who do otherwise will be left behind.”
