LexisNexis Canada President: Lawyers' Common AI Mistakes

• Samuel Puchala, president of LexisNexis Canada, identifies two common AI mistakes by lawyers: underutilization and over-reliance.
• He emphasizes that maintaining professional judgment is crucial when integrating AI tools into legal practice.
• The rapid pace of AI development makes traditional, lengthy evaluation processes for new tools obsolete.
• LexisNexis addresses lawyers' primary concern about AI accuracy by grounding its tools in verified Canadian legal content to prevent hallucinations.
• A LexisNexis Canada survey found 93% of 610 lawyers are aware of generative AI, with adoption rates accelerating.

He stresses that maintaining professional judgment remains paramount for lawyers integrating these powerful tools into their practice.

Samuel Puchala, who serves as the president and general manager of LexisNexis Canada, has identified critical missteps lawyers frequently make when engaging with artificial intelligence tools. He observes two primary errors: some practitioners briefly experiment with a new tool, encounter an unsatisfactory outcome, and subsequently abandon it; others, conversely, embrace AI, experience initial impressive results, and then place undue trust in the output. Puchala advocates for a balanced approach, emphasizing that a "happy middle ground" exists where significant efficiency and proficiency gains are achievable through these technologies. However, he stresses that maintaining professional judgment remains paramount for lawyers integrating these powerful tools into their practice. Puchala is slated to share his insights on establishing reliable and repeatable AI workflows at the upcoming Canadian Legal Summit.

His unique career trajectory, which includes studying pharmacology and toxicology at the University of Toronto, attending law school at Western University, and completing an MBA at the Ivey Business School, followed by stints at Bain & Company, Merck, and 16 years as a partner at Satov Consultants before joining LexisNexis in December 2020, provides him with a broad perspective on the intersection of technology, business, and law. This background informs his understanding of the Canadian legal AI adoption pitfalls that lawyers face today.

The legal technology landscape has undergone a profound transformation, particularly since late 2022 when large language models became publicly accessible. Samuel Puchala recalls his immediate realization upon opening his first ChatGPT window that the world would "change very quickly" for the industry. This shift prompted LexisNexis, which had been working with AI prior to this public moment, to reposition itself from a "content-led set of online tools for lawyers" to an "AI technology-first business." While its strategic direction has evolved, the company's foundational content, including legal books and its extensive case law database, continues to anchor its AI products and competitive standing.

This rapid pace of development, however, presents a new challenge for law firm buyers. Traditional sequential evaluation cycles are now obsolete, as firms that dedicate months to structured pilot programs risk making decisions based on tools that have already advanced several generations beyond their initial assessment. A LexisNexis Canada survey of 610 lawyers revealed that 93 percent are already aware of generative AI, a figure Puchala believes is now likely universal, with adoption rates accelerating year over year, consistent with coverage from Canadian Lawyer. This underscores the urgency for legal professionals to adapt their evaluation processes to the fast-evolving generative AI legal practice in Canada.

A primary concern frequently voiced by lawyers regarding AI is the accuracy of its output. Samuel Puchala distinguishes between two distinct problems often conflated under the umbrella of accuracy. The first is outright hallucination, characterized by fabricated citations or the invention of legal authorities that lack any basis in actual law. The second, more subtle issue, arises when "a real and accurate citation that’s not a hallucination doesn’t automatically mean that analysis is c."

LexisNexis has implemented robust safeguards to address these concerns, particularly within its own platform. Its AI is meticulously grounded in verified Canadian legal content, and all citations are directly linked back to their original source material. Furthermore, human expertise is integrated into the review process, significantly reducing the likelihood of hallucinations. Puchala asserts that within the LexisNexis ecosystem, there is "very little chance of a hallucination because everything is cited back, is linked back to the actual content," underscoring the company's commitment to delivering reliable and trustworthy LexisNexis AI hallucination solutions for legal professionals.

Lawyers and compliance officers should understand the common pitfalls of AI adoption – both underutilization and over-reliance – to effectively integrate these tools while maintaining professional judgment. This article highlights the importance of evaluating AI tools for grounded content to mitigate hallucination risks and adapting evaluation processes to the rapid pace of AI development.

Source: Insights derived from a recent interview with LexisNexis Canada's president.