[{"description":"CALL member Erica Friesen, Research and Instruction Librarian & Online Learning Specialist at the Lederman Law Library at Queen's University in Kingston, Ontario recently wrote an article on \" Thoughts on the New Lexis+ Brief Analysis Tool: For Law Students and Novice Researchers \". It originally appeared on Slaw.ca on September 28, 2022. It is republished here with permission of the author. New academic year; new legal research tools. Something new always comes out right as another cohort of students is gearing up to begin their law degree. And, as with many new product launches these days, “artificial intelligence” is often a prominently displayed term with accompanying materials. As legal publishers continue to launch AI-driven research tools in Canada, what do students and other novice researchers need to know to be prepared for their first forays into legal research? Lexis recently launched the latest version of their legal research platform, Lexis+ Canada, for Canadian law schools. It features significant updates that employ artificial intelligence, including a “Brief Analysis” tool in line with similar products offered by Westlaw Edge (on the US platform) and vLex (already available in Canada). According to Lexis+, this tool helps researchers “expedite completion of briefs and other legal documents.” Once your document is uploaded, it spits out a report including the following categories: 1) recommended cases, 2) cases and legislation cited in your document, 3) jurisdiction, and 4) extracted and recommended concepts. While I do not intend to comment on the utility of this tool for practitioners, I have no doubt that students beginning law school or working in summer or articling positions are bound to encounter this new feature sooner or later. When you do, you may have questions: should I be using this tool? How or how not? The landing page for this tool in Lexis+ contains no information on how this tool works and the Lexis+ Canada “Help” database so far contains minimal documentation. Confusion or uncertainty is understandable with such opacity. To gain some insight, let’s run an experiment: take two case comments written by different authors on the same Supreme Court case from 2004. [1] An older decision means that we should retrieve an abundance of recommendations from the intervening 18 years. Both comments include an overview of the area of law at the time of this decision as well as an analysis of how this particular decision would change the area of law. Once Lexis analyzed these documents, I exported the resulting report and compared the results for each comment. Here are the two documents, in case you’d like to run the experiment yourself: Comment A: Teresa Scassa, “Recalibrating Copyright Law?: A Comment on the Supreme Court of Canada’s Decision in CCH Canadian Limited et al. v. Law Society of Upper Canada ”, Case Comment, (2004) 3:2 CJLT 89. Comment B: WL Hayhurst, “The Canadian Supreme Court on Copyright: CCH Canadian Ltd. v. Law Society of Upper Canada ”, Case Comment, (2004-2005) 41 Can Bus LJ 134. Recommended Cases “Recommendations” is the section of the Brief Analysis report that lists additional case law for you to consider adding to your legal document. Brief Analysis recommended 58 unique cases in total between Comment A and Comment B. However, only 10 of those cases were the same between the two documents (after adjusting for “recommended” cases that were actually cited in the other document). See Figure 1 . Key takeaway: Legal research and writing is an art, not a science. AI-driven tools like these run on the material that you upload, so you can expect substantially different results depending on the quality of analysis and the decisions you made in producing the original document. This does not mean that there is no value in using the tool, but you cannot expect it to help you build something from nothing. Recommendations are based on the content in the document,","titre":"CALL Member Article: Thoughts on the New Lexis+ Brief Analysis Tool: For Law Students and Novice Researchers","pubDate":"Mon, 26 Oct 2026 07:55:49 +0100","id":3241473,"source":"Canadian Association of Law Libraries / L'Association canadienne des bibliothèques de droit CALL/ACBD Blog","url":"https://www.callacbd.ca/CALL-Blog/12957419"},{"description":"This article belongs to the debate » Thinking Machines, Thinking Lawyers14 October 2026The (Currently) Uneven Intelligence of AI in Comparative Constitutional Law From Years of Research to a Single Prompt? Artificial intelligence (AI) has the potential to accelerate and improve the quality of research. This potential has inspired numerous projects exploring how AI could help scholarship on constitutional law. For example, there is research using AI to interpret constitutions,1) apply constitutional principles,2) and predict court rulings.3) Even the question of whether AI could improve the process of constitution-making has been studied.4) AI’s potential is particularly significant for comparative constitutional research. Systematically comparing legal systems requires locating sources from different jurisdictions, which are often difficult to access and only available in local languages. The creation of comparative legal datasets has thus involved considerable personnel, time, and financial resources. AI can overcome many of these obstacles. AI models can search for and translate foreign-language sources, filter them, and convert the filtered information into a dataset. In doing so, AI offers the prospect of datasets — which previously required years of research and the efforts of numerous experts and research assistants to compile — being just a prompt away. However, whether AI models can generate comparative law datasets that are qualitatively comparable to human-coded datasets has not yet been addressed (although Guha et al provide an example for a broader legal benchmarking exercise).5) We therefore conducted our own benchmarking study.6) We found evidence suggesting that recent AI models may be able to code simple factual data, but they are likely unreliable substitutes for expert human coders when conceptual or interpretive judgment is required. The Benchmarking Project: Seven Coding Tasks, Four Datasets, Two LLMs In our project, we tested the ability of two large language models (LLMs) to code different types of information relevant to comparative constitutional law. We specifically assessed whether LLMs can reproduce information from four existing comparative constitutional law datasets, all of which were compiled through human coding and which served as a ‘ground truth’ for our exercise. From these datasets, we asked the LLMs to code seven variables, which can be thought of as seven tasks relevant to the creation of comparative constitutional law datasets. Each of these tasks is either a test of fact-finding or reasoning abilities. First, from the Executive Term Limit Evasion Dataset,7) we drew information on whether presidents in 106 presidential systems have evaded constitutionally mandated term limits between 2000 and 2019. The dataset includes one variable on whether presidents attempt to overstay past the end of their term. The second variable captures what strategies executives use for doing so, such as amending the constitution, replacing it with a new one, or asking the court to interpret term limits out of the constitution. While determining whether a president overstayed is a fact-finding task, assessing the strategy used to do so requires reasoning. Second, from the Comparative Constitution Making Dataset,8) we drew information on the process by which constitutions were adopted in democracies between 1900 and 2015. This dataset also includes fact-finding variables such as (1) whether citizens were involved in the formulation, discussion, or submission of reform proposals, and (2) whether citizens participated in popular referendums at any stage in the process. The dataset’s most important variable (3) captures whether constitutions reflect ‘plural elite agreement,’ that is, whether all relevant elite factions in society were involved and represented during the process. The recreation of this variable requires reasoning. Third, from the Characteristics of National Constitutions Dataset9), we drew on its coding of 43 first-generation constitutional","id":3218421,"pubDate":"Thu, 15 Oct 2026 08:26:22 +0200","titre":"The (Currently) Uneven Intelligence of AI in Comparative Constitutional Law","url":"https://verfassungsblog.de/the-currently-uneven-intelligence-of-ai-in-comparative-constitutional-law/","source":"Scoop.it - Documentation juridique (Stéphan Cottin)"}]