Artificial Intelligence: AI Tools

GAI Tools in Legal Education

GAI tools are transforming how law students study, research, and prepare for practice. These tools use advanced machine learning models to generate content—including text, images, audio, and video—based on user input. In legal education, they support everything from drafting case briefs to summarizing complex rulings and organizing research materials.

Emerging Trends for Legal Education

Agentic AI: Tools like Harvey and LawDroid Copilot are evolving into autonomous agents capable of executing multi-step legal tasks, such as drafting, reviewing, and refining documents with minimal user input.

Privacy-First AI: With growing concerns about data confidentiality, especially in legal contexts, tools like Elephas and AI Lawyer emphasize offline functionality, encryption, and user-controlled data storage.

Visual Legal Research: New platforms are emerging that use AI to visualize legal concepts, citation networks, and case relationships—enhancing comprehension and pattern recognition for complex legal topics.

AI Literacy & Ethics Integration: Law schools are increasingly embedding AI literacy and ethical use into their curricula, helping students critically assess AI outputs and understand the legal implications of using these tools.

Productivity & Study Aids

AI-Powered Summarizers

• ScholarAI: Integrates with academic databases to summarize scholarly articles, including legal journals. Great for quickly digesting dense legal theory or case law.
• Elicit: Uses AI to assist with literature reviews by extracting key points, arguments, and citations from academic texts. Ideal for legal research papers.
• Humata: Allows students to upload legal documents (e.g., cases, contracts) and ask questions about them. It generates summaries, highlights key clauses, and explains legal language in plain English.

Flashcard Generators & Note Organizers

• Quizlet AI: Automatically generates flashcards from uploaded legal texts or notes. Useful for memorizing case names, legal doctrines, and procedural rules.
• Notion AI: Enhances note-taking by summarizing lectures, organizing case briefs, and generating study checklists. It can also create timelines for legal history or procedural steps.

Citation & Formatting Tools

• Juris-M: A legal citation manager built on Zotero, tailored for Bluebook, OSCOLA, and other legal citation styles. Supports multilingual legal research.
• Zotero: A general-purpose reference manager that integrates with legal databases and word processors. Useful for organizing sources and generating bibliographies.
• Scrivener: A writing tool ideal for long-form legal writing like law review articles or theses. Helps structure arguments and manage large volumes of research.
• PerfectIt: A proofreading tool that checks for consistency in legal documents, including defined terms, abbreviations, and formatting—especially useful for law review submissions or appellate briefs.

Grammar & Style Checkers

• Grammarly: Offers grammar, punctuation, and tone suggestions. The premium version includes clarity and conciseness improvements, which are helpful for legal writing.
• Wordtune: Rewrites sentences for clarity and tone. Useful for refining legal arguments or simplifying complex language.
• Hemingway Editor: Highlights overly complex sentences and passive voice. Helps law students write more clearly and persuasively.

AI Writing Assistants

• ChatGPT: Assists with drafting legal memos, summarizing cases, and brainstorming arguments. Can simulate opposing counsel or help outline legal essays.
• Claude: Known for its thoughtful and safe responses, Claude is excellent for summarizing legal documents and generating structured outlines.
• Copilot (Microsoft): Embedded in Word and Outlook, it helps draft, summarize, and edit legal documents directly within familiar tools.

Presentation & Visual Aids

• Beautiful.ai: Automatically designs professional-looking slides. Great for moot court presentations or class lectures.
• Tome: Uses AI to generate narrative-driven presentations with embedded visuals and text. Ideal for storytelling in legal contexts (e.g., case timelines).
• Canva Magic Design: Offers AI-assisted slide creation with legal-themed templates. Useful for creating infographics, flowcharts, or visual case briefs.

RAG tools leverage the strengths of LLMs by integrating them with external knowledge repositories. While LLMs are designed to understand and generate human language with remarkable fluency, RAG takes their capabilities a step further by enabling them to tap into more precise and relevant sources of information, such as specialized documents, extensive databases, and online resources.

Some of the notable features of RAG tools are as follows:

Factual Accuracy: By utilizing up-to-date external knowledge sources, RAG tools are able to provide users with more reliable information that is grounded in current facts, reducing the likelihood of errors that may arise from the model's inherent limitations in knowledge.

Contextual Relevance: RAG tools are designed to deliver responses that are not only accurate but also highly relevant to the specific context of user inquiries. By interpreting the nuances of a question, these tools can generate tailored responses that meet the unique needs and situations of users.

Improved Performance: The integration of external knowledge with LLMs leads to significant enhancements in the overall quality and consistency of outputs. This means that users can expect more coherent and context-aware responses, which enhances the user experience and makes interactions more productive.

GAI and RAG tools signify major technological progress. GAI tools boost human creativity and productivity, while RAG tools improve the reliability and informative value of AI systems.

Focus

• GAI tools primarily focus on creative content generation.
• RAG tools prioritize factual accuracy and information retrieval.

Data Sources

• GAI tools primarily rely on internal training data.
• RAG tools integrate external knowledge sources.   

Applications

• GAI tools have a broader range of applications.
• RAG tools are particularly well-suited for tasks that require access to specific information.

Agentic AI refers to autonomous artificial intelligence systems capable of achieving goals without constant human intervention. Key capabilities include:

Autonomous Decision-Making: These systems analyze situations and make independent decisions.

Goal-Oriented Behavior: They break down complex tasks into smaller steps to efficiently reach objectives.

Proactive Actions: Agentic AI identifies opportunities and potential issues, taking initiative instead of merely reacting.

Continuous Learning: These systems improve over time by learning from experiences.

An example is a self-driving car, which constantly analyzes traffic and makes real-time decisions for safe navigation. Agentic AI represents a shift from passive systems to proactive, autonomous ones.

OpenAI’s Deep Research is an agentic AI feature in ChatGPT that allows users to autonomously browse the web and analyze sources for thorough research. It can plan and execute tasks, generating detailed, citation-backed reports, which are beneficial for legal, academic, and policy work. While it enhances research efficiency and accuracy, it occasionally misinterprets complex material or produces errors.

AI encompasses various subfields, with GAI and agentic AI being two prominent types. GAI represents the creative aspect of AI, focusing on generating new content (like text, images, or music) by analyzing and mimicking patterns in existing data. This is evident in tools like ChatGPT and DALL-E. 

On the other hand, agentic AI is designed to operate autonomously in order to achieve specific objectives. It is characterized by its ability to make decisions and respond dynamically to real-time information, as seen in applications like autonomous vehicles and robotic process automation. For example, a smart assistant managing your calendar showcases the capabilities of agentic AI.

While GAI typically works within predefined boundaries, agentic AI is more flexible and responsive. These two types of AI can complement one another effectively; for instance, in a virtual customer service agent, GAI can craft personalized responses while agentic AI manages the interaction in real-time.

For more information, check out Bernard Marr's excellent article Generative AI Vs. Agentic AI: The Key Differences Everyone Needs To Know.

GAI in Legal Research

GAI tools can assist with drafting legal memos, summarizing case law, and explaining complex legal concepts. However, they come with important limitations, including potential inaccuracies, biases in training data, and ethical concerns around confidentiality and attribution. A particularly serious issue is AI hallucination—when the model generates plausible-sounding but false information, such as nonexistent case citations. GAI tools do not inherently verify facts or cite sources, which makes human oversight essential. Law students and professionals must critically evaluate GAI-generated content and cross-check it against authoritative legal sources.

Examples: ChatGPT, Claude, Gemini, Copilot

Agentic AI in Legal Research

Agentic AI represents the next evolution in legal research tools. These systems can autonomously plan and execute multi-step tasks, such as conducting research, drafting documents, and formatting citations—all with minimal user input. They simulate the behavior of a legal assistant or junior associate, making them especially useful for complex workflows.

Examples:

Harvey AI: Built on OpenAI’s models, Harvey can handle end-to-end legal tasks like contract review and compliance analysis.

LawDroid Copilot: A virtual legal assistant that automates document drafting and client intake.

ChatGPT’s Deep Research: An agentic feature that autonomously browses the web, analyzes sources, and generates detailed, citation-backed reports. It is particularly useful for legal research and policy analysis, though it occasionally misinterprets complex material.

RAG Legal Research Tools

RAG tools offer a significant advancement over standalone GAI by combining generative capabilities with real-time retrieval from trusted legal databases. This hybrid approach enhances factual accuracy, reduces hallucinations, and improves transparency by providing source citations. RAG tools are often tailored for the legal domain, enabling them to interpret legal terminology and context more effectively.

Examples:

Lexis+ AI & Westlaw Precision AI: These platforms retrieve and synthesize legal information from proprietary databases, offering grounded responses with citations.

Perplexity AI: While not legal-specific, it provides cited answers and is increasingly used for preliminary legal research.

Key Features:

Database-Driven: Pulls from case law, statutes, and regulations.

Contextual Generation: Drafts summaries and analyses based on retrieved legal content.

Transparency: Includes citations for verification and auditability.

Emerging Trends

Agentic Workflows: Tools like Deep Research and Harvey are moving beyond simple Q&A to perform full research and drafting tasks.

Legal-Specific Fine-Tuning: AI models are increasingly being trained on legal corpora to improve accuracy and relevance.

Ethical AI Use: Law schools are embedding AI literacy and ethics into curricula, emphasizing responsible use, data privacy, and academic integrity.

Hybrid Human-AI Collaboration: AI is being positioned as a co-pilot, not a replacement—supporting, not supplanting, legal reasoning.

Westlaw's homepage will take users to AI-Assisted Research. For conventional legal research, select the "Precision Research" tab.

Westlaw Precision with CoCounsel is the most advanced version of Westlaw, incorporating GAI capabilities to enhance and streamline legal research. CoCounsel offers several key GAI features:

AI-Assisted Research: Provides quick answers to legal research questions.

Claims Explorer: Identifies the strongest claims related to a matter and can reveal additional relevant claims.

AI Jurisdictional Surveys: Offers insights into legal topics across different jurisdictions.

Quick Check: Rapidly detects mischaracterizations of cited authority within uploaded documents.

Use these tools to expedite research, but do not rely on them as a substitute for a comprehensive investigation.

Westlaw's Practical Law is a legal research and knowledge platform designed to support lawyers in their daily practice. It is organized by practice area, sector, resource type, and jurisdiction. Practical Law goes beyond traditional legal research by offering a wealth of practical guidance and tools, including the following features:

Standard Documents: Access to a comprehensive library of customizable legal documents, including contracts, agreements, and pleadings, tailored to meet specific client needs.  

Checklists: Detailed checklists encompass various legal procedures and transactions to ensure that lawyers do not overlook essential steps.

How-to Guides: Clear, step-by-step instructions and practical advice on various legal topics, including drafting contracts and conducting discovery. 

Legal Updates: Timely updates on legal developments, such as new legislation, court decisions, and regulatory changes.   

Expert Guidance: Content created and maintained by experienced legal professionals, ensuring reliable and up-to-date information.

AI-Assisted Research utilizes LLMs to analyze the language found in cases, statutes, and other primary legal authorities to enhance accuracy. The responses reference primary law using the actual wording from these sources, and links are provided for accessing complete primary law documents. However, despite these efforts, AI-Assisted Research may occasionally produce inaccuracies. Therefore, it should always be used as part of a broader research process, supplemented by additional research to fully understand the nuances of the issues and improve accuracy.

The AI-generated summary of results, located above the list of primary legal authorities, can be very helpful for gaining an overview of the issues and identifying relevant primary authorities. However, it should never be used to advise a client, draft a brief or motion for a court, or be relied upon without conducting further research.

To use AI-Assisted Research, formulate a legal research question as you would naturally express it to a colleague. Write a concise and focused query that is one or two sentences long. Include relevant details such as the practice area, cause of action, governing law, remedy sought, desired outcome, and any pertinent material facts.

Claims Explorer allows users to input facts that may lead to a potential legal claim. It provides results for disputes between private parties or claims against government entities that might involve damages. Claims Explorer utilizes LLMs and incorporates Westlaw's editorial content to suggest potential claims based on the provided factual information. These suggestions are based on the most recent versions of statutes, and reviewing past statutory language can help verify that recent amendments do not affect a potential claim. Claims Explorer can search all claims, as well as state claims from California, Florida, Illinois, New York, Pennsylvania, and Texas.

While Claims Explorer can be valuable for gaining an overview of issues and identifying primary sources of authority, it is important to note that it may sometimes produce inaccuracies. Therefore, it should always be used as part of a broader research process. The AI-generated list of potential claims should not be used to advise clients, draft complaints, or be relied upon without conducting further research.

For Claims Explorer, follow these suggestions to build effective queries:

• Draft a query that is approximately three or four sentences in length.
• Describe the fact pattern in general language.
• Include the types of parties involved rather than the names of the parties.

AI Jurisdictional Surveys can be used to analyze, identify, and compare statutes and regulations on a specific topic. As mentioned above, AI Jurisdictional Surveys use LLMs and can produce inaccuracies, so always use the survey results alongside additional primary source research.

Follow these suggestions when using AI jurisdictional surveys:

• Compose a clear, concise, and focused query that can be answered by statutes and regulations.
• Avoid writing in a prompt, command, or instruction style.
• Do not include jurisdictions or terms that indicate jurisdictions in the query.
• Concentrate on a single issue or question, and refrain from including proper names of individuals, parties, or locations.

Quick Check leverages AI to go beyond traditional legal research methodologies and recommend relevant authorities that may have been missed. Quick Check examines case quotes in context, providing alerts and explanations of potential mischaracterizations. With Quick Check, users can upload documents to check their work, analyze an opponent's work, and see analyses of briefs for both parties.

• Check Your Work: This feature allows users to review early drafts to expedite research, verify completeness before filing, and update outdated briefs with current authority.
• Analyze an Opponent's Work: This function helps identify issues with opposing citations, find contrary authority, and analyze one's own work to anticipate criticisms or weaknesses.
• Quick Check Judicial: This option offers analysis similar to what judges see, reviews relevant authority not cited by either party, and identifies issues with the citations and quotations used.

Pertinent security measures are included on the Quick Check page.

On the Lexis homepage, select Protégé to ask a legal question, generate a draft, summarize a case, or upload documents.

• Ask legal questions conversationally and receive answers with citations and links to LexisNexis content that you can verify and use in your legal research.
• Draft legal arguments, research notes, letters, emails, and clauses.
• Generate a summary of a case, giving an overview of the case, legal reasoning, and its outcome.
• Upload your documents and ask questions about those documents or have Lexis generate a summary of the uploaded documents.

The Lexis Protégé product uses “agentic AI” – a technology that can autonomously complete tasks based on user goals, including reviewing its work and identifying improvement areas. It can draft documents, analyze legal data, conduct research, and provide workflow suggestions, significantly improving efficiency and accuracy for legal professionals. Key features include document drafting, legal research, workflow automation, and data analysis. The system can process significantly larger documents than previous versions, handling up to one million characters or approximately 300 pages.

Use these tools to expedite research, but do not rely on them as a substitute for a comprehensive investigation.

Practical Guidance by LexisNexis offers legal professionals comprehensive tools to streamline their practice. It provides step-by-step instructions, checklists, and templates to facilitate efficient legal research and drafting. Users benefit from expert insights that help navigate complex legal issues across various practice areas. The platform is designed to enhance productivity, allowing attorneys to focus more on their clients and less on administrative tasks. Additionally, it features an extensive library of resources that are continually updated to reflect the latest legal developments. Overall, Practical Guidance serves as an invaluable resource for attorneys looking to improve their workflow and deliver high-quality legal services.

Use the Ask a Legal Question (Ask) task to inquire about a legal topic or issue, such as finding a specific case or legislation. You can ask about the following:

• Legal definitions
• Current law questions
• Procedural inquiries
• More complex legal scenarios  

You can enter up to ten prompts in each conversation. As you submit additional prompts, Lexis evaluates them alongside previous prompts and responses. After receiving the initial reply, Lexis+ AI will present some sample follow-up prompts above the "Ask a Legal Question" field. You may enter up to nine follow-up questions after your initial prompt.

Tips for Creating Prompts 

In Lexis, a prompt is akin to asking questions or providing information during a conversation. For the best results, consider the following tips when formulating prompts:

• Clearly state your goal, e.g., “Draft a letter to my client regarding a retainer agreement.”
• Include any parties, key terms, or legal points relevant to your inquiry.
• Provide material facts and necessary context for a thorough response.
• Specify the type and format of the answer you expect, e.g., “Find statutes in Illinois regarding speed limits on unrestricted highways.”
• Refine your search by submitting follow-up prompts to narrow or expand your query, e.g., “Find more cases like this in Indiana.”

If the response to your first prompt does not meet your expectations, ask a follow-up question that includes additional information.

With the Generate a Draft (Draft) feature, users can draft the following:

• Legal Arguments
• Research Notes
• Letters
• Emails
• Clauses

Please note that there are examples of the types of prompts that can be used to create a successful draft. Additionally, users can alter the tone of voice in the generated draft. For instance, the following instructions can be entered in the input bar:

• Shorten and simplify.
• Soften tone or make more assertive.
• Explain what changed.

To create the best draft response, follow these suggestions:

• Specify the type of draft you need, such as "Draft a client email explaining the circumstances under which a non-compete agreement will be enforceable against a non-executive-level employee."
• Provide sufficient context and background to help Lexis understand what you want to generate.
• If you do not receive the draft you are seeking, ask a follow-up question with more details.

Use the Summarize a Case (Summarize) task to create an overview of a specific court case's facts, issues, and legal reasoning.

• A summary provides an overview of a case, its legal reasoning, and its outcome.
• Lexis responds best to full citation information (the more specific the citation, the better the response will be).
• Generating a summary of a case using Lexis simplifies the search experience by reducing the need to read through a large case report. 
• The typical summary consists of headings and bullet points describing the case's key aspects.

The Upload to Summarize or Ask a Question (Document) task extracts and summarizes key insights from selected documents. Users can upload ten files, with a maximum size of 20MB per file, during a single session.

Once your documents are uploaded, the bottom buttons are active.

• Click on the "ask questions about your uploads" button to ask questions about your documents, e.g., ask Lexis to find the main arguments in the documents.
• Click on the "summarize your uploads" button to have Lexis generate a summary of your uploaded documents.

NOTE: All files are encrypted when uploaded, saved temporarily, and deleted at the end of each session.

TIP: Multiple documents can be selected at once by either dragging and dropping or using the browse function. Hold down the CTRL and SHIFT keys as you would in a file explorer to select them.