A New Brunswick lawyer might face disciplinary motion after allegedly utilizing generative AI improperly to write down a authorized temporary, and it’s sparking conversations about the usage of AI within the occupation.
The Legislation Society of New Brunswick alleged that in April 2025, Denis Sawyer submitted a quick through which 10 of 12 citations had been fictitious.
“When Mr. Sawyer was knowledgeable of this, he filed a subsequent temporary, through which he eliminated nearly all fictitious citations besides for 2,” a public discover states.
Sawyer filed a response in July stating he admitted to the contents of the discover and that accepts “full duty for my oversight.”
However because the matter is about for a listening to subsequent month, authorized specialists say the difficulty shouldn’t be if legal professionals are utilizing AI, however quite how they’re utilizing it.
“Utilizing AI shouldn’t be unethical in itself. What’s unethical is utilizing AI incompetently,” Norman Siebrasse, a professor within the college of regulation on the College of New Brunswick, instructed International Information.
Get day by day Nationwide information
Get day by day Canada information delivered to your inbox so you may by no means miss the day’s high tales.
Siebrasse stated specialised AI instruments are already getting used to assessment paperwork, analyze contracts and enhance effectivity.
“AI actually does have the potential to enhance authorized apply and make authorized recommendation extra reasonably priced and higher,” he stated.
However others say these advantages can solely happen if legal professionals do their due diligence to make sure the accuracy of the knowledge.
“When a lawyer is submitting materials to courtroom, the buck stops with that lawyer, whether or not or not they obtain assist from a human or a know-how instrument,” Amy Salyzyn, a professor within the college of regulation on the College of Ottawa, stated.
However errors have gotten extra frequent.
Seven choices involving AI-fabricated authorized supplies had been recorded by Canadian courts and tribunals in 2024, in accordance with knowledge from Courtyready, a justice watchdog. By 2025, that quantity had risen to 87 choices. By the top of the primary half of 2026, 74 such circumstances had been documented.
Karen Etis, a regulation professor on the College of Ottawa, instructed International Information that the dangers of utilizing AI in regulation transcend citing incorrect data. She stated one other concern is how these studying fashions can set precedents.
“You’re destroying the reliability of precedent,” she stated. “As a result of it’s attainable to set a case that another person will then cite and another person will then construct on that and it’s now not dependable.”
Etis additionally expressed issues about with how AI can change the best way legal professionals deal with confidential consumer data.
“These items are free as a result of they prepare in your knowledge,” she stated.
Legislation faculties are already making ready to show the following technology of legal professionals tips on how to use AI responsibly because the know-how turns into more and more prevalent in authorized apply.
Siebrasse, from the College of New Brunswick, stated correct schooling may help future legal professionals use these instruments to their profit.
“We’re taking steps to organize our college students for AI-assisted apply and in addition to combine AI into our pedagogy,” he stated.
Sawyer’s disciplinary matter can be heard subsequent month.
© 2026 International Information, a division of Corus Leisure Inc.
