A stack of documents in an arbitration room once signalled days of painstaking review by junior associates. Today, artificial intelligence quietly sifts through hundreds of thousands of pages, highlighting key passages, summarizing witness interviews, and even flagging relevant precedents. The future of legal practice is arriving faster than many lawyers expected – and not just in the courtroom.
Tracey Cohen, a commercial litigation and arbitration counsel partner with Fasken Martineau DuMoulin LLP, sees AI as a “potentially extremely useful” tool that can dramatically improve efficiency and effectiveness when used appropriately.
“From transcription and translation to synthesizing large volumes of documents, creating chronologies, or turning what might be written submissions into a PowerPoint format – all these tools are helpful,” the Vancouver-based Cohen notes. Beyond document management, Cohen highlights AI’s potential role in proofreading memorials, supporting legal research, and helping select arbitrators or mediators based on skill sets and prior related experience.
However, she cautions that predictive analysis remains limited in commercial arbitration because most proceedings are confidential, making it difficult to amass the large data sets needed for effective AI analysis. While public court decisions offer richer data sets, the private nature of commercial arbitration presents unique challenges.
She says that accuracy, bias, privacy, privilege, cybersecurity, and enforceability are central concerns. Improper AI use could jeopardize an award and its enforceability. Cohen notes that guidelines from law societies, courts, and arbitral institutions emphasize fairness, due process, independence, and transparency. Proper training of lawyers and staff is essential to maximizing AI’s benefits.
“Learning how to prompt effectively is fascinating… Teaching people how to speak to these tools and ensuring that the tool is only pulling from reliable data sets is part of what we’re spending a lot of time figuring out.”
Looking ahead, Cohen anticipates AI will be used to handle low-value, straightforward disputes. She notes that the American Arbitration Association–International Centre for Dispute Resolution (AAA-ICDR) has announced an AI-based arbitrator tool to deliver fast, cost-effective, and trusted dispute resolution for low-value construction disputes, starting in November 2025.
Says Cohen: “This is being tested where the decision is based upon documentary evidence, where there are no witness credibility issues, and the amount at stake likely doesn’t warrant traditional arbitration or litigation.”
Adam Goldenberg, a partner in the litigation group at McCarthy Tétrault LLP who represents Canadian and global companies in commercial and public law disputes, says his firm sees “enormous potential for artificial intelligence to enhance efficiency and outcomes in litigation.”
AI tools are already integral to the firm’s disputes practice, especially within its MT3 division, which builds bespoke AI models to expedite review in complex investigations and litigation matters. “This enables lawyers to focus on tasks requiring human judgment, and that really adds value for clients.”
Goldenberg also notes AI’s role in discovery and document review, from identifying key documents and building chronologies to generating outlines for witness examinations and agreed statements of fact. He says combining AI tools with legal expertise and oversight allows for the rapid analysis of vast data sources, and projects that previously took months are now completed in a fraction of the time.
Goldenberg says McCarthy Tétrault uses a combination of off-the-shelf and proprietary AI systems, all tailored to strict confidentiality and privilege standards, with most tools operating on the firm’s servers rather than the cloud. Goldenberg warns, “You should never, as a lawyer, be sharing confidential information with a public website.”
Human oversight is essential, Goldenberg says. “AI is not, and will not, replace the role of human lawyers and decision-makers in ADR processes.” Decision-making and responsibility must remain with humans.
He also observes that courts and arbitral institutions are establishing AI policies, requiring disclosure when AI is used, and that case law has begun imposing sanctions for the improper use of AI.
While lawyers continue to rely on court decisions and case law, Goldenberg sees AI assisting in outcome prediction, especially in specialized areas like tax or immigration. He believes AI will make lawyers better researchers and writers and foresees a shift toward value-based billing.
Mike Mestinsek, head of the advocacy group in the Calgary office of Stikeman Elliott LLP, echoes the transformative potential of AI in dispute resolution. He notes that what began as a crude tool for e-discovery has evolved into a dependable necessity.
“It’s just a reality now,” Mestinsek says. AI now efficiently manages document production with safeguards like quality-control sampling and clawback provisions to mitigate errors.
Client demand is also driving the adoption of AI as an efficiency and potential cost-saving tool. Clients “are not going to pay for old models of work if they feel AI can do the job faster and just as thoroughly, or even better,” he observes. Mestinsek notes that even traditionally cautious firms are accelerating implementation to meet client expectations.
Firms are responding by creating secure, closed AI environments to protect client information.
“You can’t just jump on ChatGPT,” he cautions. Mestinsek describes the profession’s current approach as controlled experimentation, with larger firms building proprietary platforms and smaller firms relying on vetted third-party software.
Mestinsek emphasizes that while AI has transformed repetitive legal work, it cannot replace decision-making, which remains the lawyer’s most valuable skill. “AI can give you data and analytics, but it’s no substitute for judgment,” he says. Strategic advice, ethical reasoning, and understanding client goals will always require human insight.
He also cautions that overreliance on AI could erode foundational training. Junior lawyers who no longer spend hours on document review may miss key learning experiences that build expertise and writing ability. “If AI is doing most of the writing, you’re not developing as a writer,” he says. Maintaining training opportunities will be essential as AI evolves.
For lawyers, the challenge is to integrate technology wisely, leveraging efficiency without losing the human judgment that defines good lawyering.
Vasilis Pappas, partner and co-head of Bennett Jones LLP’s international arbitration practice group, emphasizes AI’s rapid evolution and transformative impact on legal practice in arbitration. “It’s evolving so fast that what I say today may be completely obsolete tomorrow,” he notes. “The type of activities that armies of associates used to undertake over weeks can now be undertaken in hours by an AI platform.”
Pappas also highlights AI’s utility in international arbitration, especially in legal research, drafting, identification of arbitrators, transcriptions, document translation, and real-time interpretation at hearings. “Now you can … more efficiently translate huge volumes of documents,” he says.
Despite these benefits, Pappas cautions against overreliance on AI at this point. “These are still imperfect tools… You still need that double-checking by a human lawyer.”
While AI may boost efficiency, Pappas is cautious about whether it will ultimately reduce costs. He recalls the digitization of legal research with Westlaw and LexisNexis a few decades ago, which was assumed would reduce costs but “actually did the opposite” despite increasing productivity. Those platforms were expensive and significantly increased the availability of information requiring review. The same could occur with AI platforms.
Nevertheless, Pappas stresses the importance of preparing young lawyers to work effectively with AI: “I would rather that our young associates … learn how to use it in a way that makes sense for our practice and for our clients.”
Pappas foresees AI handling low-value disputes and notes that several institutions already offer online dispute resolution (ODR) platforms. “There is a world … where relatively simple, low-value disputes can be resolved essentially by AI.” However, he warns that complex, high-value cases require nuance, judgment, and emotional intelligence that AI may not be able to replicate soon.
Sarah McEachern, a partner at Borden Ladner Gervais LLP in Vancouver specializing in commercial and construction disputes, notes that many AI-enabled tools are already embedded in practice, particularly in e-discovery, for complex, document-heavy disputes.
She says platforms with predictive coding allow lawyers to review large amounts of data more efficiently than manually. Still, McEachern cautions “that one excellent, relevant document might not be flagged,” underscoring the continuing need for strategic human judgment.
While AI can assist with rapid legal research and case identification using verified platforms, open tools like ChatGPT may produce hallucinations – results that sound persuasive but can’t be verified. She says properly checking sources and evaluating the AI-generated material for context remains essential.
McEachern notes AI’s value in transcript review and evidence management, particularly for lengthy hearings or discovery. Generative tools can quickly locate references, highlight admissions, or identify evidentiary gaps, freeing junior lawyers to focus on substantive strategy. However, she acknowledges that traditional research tasks also cultivate analytical skills.
She also points to broader AI developments in online dispute resolution (ODR), including platforms like PayPal and eBay and mediation tools that use algorithms to resolve small claims efficiently. However, she warns, AI tools lack “nuance, emotional intelligence, fairness, and contextual understanding,” all critical in higher-stakes arbitration.
From the perspective of arbitrators, McEachern sees AI streamlining administrative tasks – such as drafting procedural histories or summarizing cross-examination transcripts – if parties give consent and all decision-making authority remains with the tribunal.
Finally, the BLG partner stresses that arbitrators and counsel must stay alert to emerging AI guidelines from courts and arbitration institutions such as the Chartered Institute of Arbitrators and the ICDR’s Guidance on Arbitrators’ Use of AI Tools. McEachern also notes that while details of arbitration rules can differ slightly in different jurisdictions, these guidelines provide valuable benchmarks for ethical and professional AI use.
From arbitration to litigation, AI is redefining how lawyers work – accelerating processes, enhancing analysis, and training future generations. However, lawyers specializing in ADR agree on one principle: AI complements human expertise but does not replace it. The challenge is striking a balance: harnessing AI’s efficiency while ensuring that lawyering remains grounded in judgment, nuance, and responsibility. AI is changing dispute resolution, but judgment matters most.
Lawyers’ use of AI for ADR
How often have you used, and do you expect to use, AI tools and technology?
Source: White & Case 2025 International Arbitration Survey