The principle is simple: Use AI as a powerful assistant, not an autopilot. Every best practice below reinforces the same core idea — AI amplifies your expertise when you stay in control of the process, verify the output, and apply your professional judgment at every step.
Verify Every AI Output Against Primary Sources
AI language models can generate convincing but inaccurate legal text, including fabricated case citations. Always verify AI-generated content against authoritative primary sources before relying on it in any professional context.
How to Do It
Cross-reference every AI-generated legal citation, statute, and factual claim with official legal databases and primary source materials.
Steps
- Run your AI query and capture the full output, including any cited cases, statutes, or regulations.
- Open a reliable legal research platform such as Westlaw, LexisNexis, or Google Scholar Case Law.
- Search for each cited case by name, citation number, and court to confirm it exists.
- Read the actual holding of each case to confirm the AI's characterization is accurate.
- Verify that statutes and regulations cited are current and have not been amended or repealed.
- Check secondary claims (dates, court names, jurisdictions) against official court records.
- Flag and remove any hallucinated or inaccurate references before incorporating the output into your work product.
- Document your verification process so you can demonstrate due diligence if challenged.
Tools
Use AI to Accelerate Research, Not Replace It
AI excels at helping lawyers find starting points, identify relevant legal concepts, and survey broad areas of law quickly. It should augment your research skills, not serve as a substitute for thorough legal analysis.
How to Do It
Use AI as a research accelerator to identify leads, generate keyword lists, and map legal landscapes, then conduct proper research using authoritative databases.
Steps
- Define your research question clearly before engaging the AI tool.
- Use AI to brainstorm relevant legal theories, doctrines, and statutory frameworks that may apply.
- Ask the AI to generate lists of search terms and Boolean queries you can use in Westlaw or LexisNexis.
- Use AI to summarize complex areas of law as a starting framework, then verify with primary sources.
- Run the AI-generated search queries in your legal research database to find actual authorities.
- Use AI to help organize and outline the results of your verified research.
- Conduct the final legal analysis yourself, applying your professional judgment to the verified materials.
Tools
Establish a Firm-Wide AI Use Policy
Every law firm and legal department should have a clear, written policy governing how AI tools may be used. This protects the firm, its clients, and its professional obligations.
How to Do It
Draft and implement a comprehensive policy covering approved tools, data handling, client confidentiality, quality control, and ethical compliance.
Steps
- Assemble a working group including partners, associates, IT, and ethics/compliance personnel.
- Audit current AI use across the firm to understand what tools are already being used and how.
- Define which AI tools are approved for use and which are prohibited.
- Establish data handling rules specifying what information may and may not be entered into AI tools.
- Create mandatory verification and review procedures for AI-generated work product.
- Define disclosure requirements for clients and courts.
- Set training requirements for all personnel who will use AI tools.
- Establish a regular review cycle to update the policy as technology and regulations evolve.
- Distribute the policy firm-wide and require written acknowledgment from all personnel.
Tools
Disclose AI Use When Required
A growing number of courts, bar associations, and regulatory bodies require or recommend disclosure of AI use in legal proceedings. Proactive disclosure protects you, your client, and the integrity of the legal system.
How to Do It
Track disclosure requirements across all jurisdictions where you practice, build disclosure into your standard workflows, and err on the side of transparency.
Steps
- Research AI disclosure requirements in every jurisdiction where you file or appear.
- Monitor court-specific standing orders, as many judges have issued individual AI-related orders.
- Update your engagement letters to address AI use and obtain client consent.
- Establish a firm-wide tracking system for jurisdictional disclosure rules.
- Include AI disclosure language in filings where required by local rule or standing order.
- When in doubt, disclose. Voluntary disclosure rarely causes harm; failure to disclose can be devastating.
- Maintain records of which AI tools were used and what verification was performed for each matter.
Tools
Use AI for First-Draft Generation
AI tools are remarkably effective at generating first drafts of legal documents, saving hours of work on routine writing tasks. The key is to use AI output as a starting point that you then refine with your professional expertise.
How to Do It
Use AI to generate initial drafts of common legal documents, then systematically review, verify, and refine them to meet professional standards.
Steps
- Identify the document type and gather all relevant facts, terms, and client-specific requirements before prompting.
- Craft a detailed prompt that includes document type, jurisdiction, key terms, parties, and any specific provisions needed.
- Provide the AI with relevant context such as template language, applicable rules, or excerpts from similar documents.
- Generate the first draft and immediately review it for structural completeness.
- Verify all legal citations, statutory references, and regulatory requirements.
- Customize the draft to reflect client-specific facts, negotiated terms, and strategic considerations.
- Review for jurisdictional accuracy, ensuring the document complies with local rules and conventions.
- Have a second attorney review the final product before delivery or filing.
Tools
Leverage AI for Document Review
AI-powered document review tools can dramatically reduce the time and cost of reviewing large document sets in litigation, due diligence, and regulatory matters, while often matching or exceeding human accuracy for certain tasks.
How to Do It
Integrate AI-powered review tools into your document review workflow for faster, more consistent, and more cost-effective review of large document collections.
Steps
- Define your review objectives clearly: what are you looking for and what are the key issues?
- Select an appropriate AI review platform based on your matter requirements and budget.
- Upload and process the document set, applying initial AI-powered culling to remove clearly non-responsive documents.
- Use AI-assisted coding to train the model on a sample set of documents reviewed by experienced attorneys.
- Run the AI classification across the full document set and review the results.
- Conduct quality control checks on a statistically significant sample of AI-classified documents.
- Have human reviewers handle edge cases, privileged documents, and any documents the AI flagged as uncertain.
- Document your review methodology for defensibility, including AI tools used and quality control measures.
Tools
Build a Personal Prompt Library
Developing a curated library of tested, effective prompts is one of the highest-leverage investments a lawyer can make in AI productivity. Good prompts consistently produce better output, and a well-organized library saves time on every future use.
How to Do It
Systematically develop, test, refine, and organize prompts for your most common legal tasks, building a reusable library that improves over time.
Steps
- Identify your 10-20 most common recurring tasks that could benefit from AI assistance.
- Draft initial prompts for each task, following structured prompting best practices.
- Test each prompt across multiple AI tools and evaluate the quality of output.
- Refine prompts based on results, adding specificity where output was too vague and constraints where output was off-target.
- Organize prompts by category (research, drafting, analysis, correspondence) in a centralized document or tool.
- Include metadata for each prompt: purpose, best AI tool to use, sample output quality, and any known limitations.
- Share effective prompts with colleagues and incorporate their feedback.
- Review and update your library monthly as AI tools evolve and your needs change.
Tools
Stay Current on AI Developments
AI technology and its regulation are evolving at an unprecedented pace. Lawyers who fall behind risk both malpractice and competitive disadvantage. Staying informed is now a component of professional competence.
How to Do It
Build a sustainable routine for tracking AI developments relevant to legal practice, including new tools, regulations, ethical guidance, and case law.
Steps
- Subscribe to 3-5 high-quality newsletters focused on AI in legal practice.
- Follow key thought leaders, bar associations, and legal technology organizations on social media or RSS feeds.
- Attend at least one AI-focused CLE program per quarter.
- Join a professional community or working group focused on legal AI.
- Set up alerts for key topics: AI regulation, legal ethics + AI, and developments in your practice area.
- Dedicate 30 minutes per week to reading about AI developments.
- Share relevant developments with your team through a regular digest or internal channel.
- Review new AI tools quarterly to assess whether they could improve your practice.
Tools
Train Your Team
AI tools are only as effective as the people using them. Investing in comprehensive, ongoing training ensures your team uses AI safely, ethically, and productively, while avoiding the pitfalls that have led to sanctions and malpractice claims.
How to Do It
Design and implement a structured AI training program that covers tool proficiency, ethical obligations, risk management, and practical application for all members of your legal team.
Steps
- Assess current AI knowledge levels across your team through a brief survey or informal conversations.
- Design a tiered training curriculum: foundational literacy, intermediate application, and advanced techniques.
- Deliver foundational training covering AI basics, firm policy, ethical obligations, and common risks.
- Provide hands-on workshops where team members practice using approved AI tools on real (or realistic) tasks.
- Create practice-area-specific training modules addressing how AI applies to each group's work.
- Establish a mentorship or buddy system pairing AI-confident team members with those who are less experienced.
- Schedule recurring training sessions (at least quarterly) to cover new tools, new risks, and lessons learned.
- Measure training effectiveness through follow-up assessments and monitor AI usage patterns.
Tools
Experiment in Low-Risk Contexts First
The safest way to build AI proficiency is to start with tasks where mistakes carry minimal consequences. Low-risk experimentation lets you learn AI capabilities and limitations before applying them to high-stakes legal work.
How to Do It
Identify low-risk, low-stakes tasks to serve as your AI learning laboratory, then gradually expand to more complex applications as your confidence and competence grow.
Steps
- Identify tasks in your practice that are low-risk: internal memos, non-binding research summaries, administrative tasks, or personal projects.
- Choose one approved AI tool and start with a single, simple use case.
- Complete the task both with AI and without AI, then compare the results for quality and time spent.
- Gradually increase the complexity and stakes of your AI-assisted tasks as you build confidence.
- Document what works, what fails, and what lessons you learn at each stage.
- Move to moderate-risk tasks only after you have developed reliable verification habits.
- Approach high-stakes tasks with AI assistance only when you have a proven workflow and strong verification discipline.
- Maintain a learning journal or log that tracks your AI experiments and outcomes.
Tools
Fundamental Best Practices
Principles every legal professional should follow when working with AI.
Verify everything against primary sources
Every citation, every statute, every factual claim must be checked against authoritative sources. Use Westlaw, LexisNexis, or the official databases of your jurisdiction.
Establish an AI policy in your firm
Define which tools are approved, for which uses, with what safeguards, and under what supervision. A clear policy protects the firm, the attorneys, and the clients. Include confidentiality and verification protocols.
Always specify your jurisdiction
When crafting prompts, state the country, state, and applicable area of law. The difference between Texas law and California law can be as significant as between English law and French law.
Start with low-risk tasks
Document summaries, exploratory research, information organization. Advance gradually to more complex tasks as you develop experience and calibrated confidence in the tool.
Protect your clients' data
Use enterprise versions with data processing agreements. Anonymize sensitive information before entering it. Comply with applicable data protection regulations in your jurisdiction.
Document your process
Keep a record of which tools you used, for what tasks, what verifications you performed, and what professional judgment you applied. If your work is questioned, this documentation demonstrates due diligence.
Ready to Put This Into Practice?
The best way to learn is by doing. Our Quick Wins give you step-by-step AI exercises you can complete in minutes, each one applying the best practices you just learned.
Ready for structured learning? Explore the Learning Program →
Comments
Loading comments...