AI for law firms: how lawyers use it and the ethics rules that apply
Short answer: Law firms use AI for legal research, first drafts, document review, summaries and client intake. ABA Formal Opinion 512, issued in July 2024, says lawyers can use generative AI if they stay competent, protect client confidences, talk to clients about it and charge fair fees. Courts are sanctioning lawyers who file AI-written briefs with fake citations, so every citation must be checked by a person before filing.
You're at a law firm and want to know what AI can actually help with, and exactly where the ethics rules and real sanctions cases draw the line.
I'm Ranjit, and I run Blur Studio: 15+ years building software, 120 apps taken live, and clients who have gone on to raise $10M+. I build AI systems and processes that help businesses do a lot more with a lot less time. I also work differently from most of what else is out there, more on that further down this page. Say hello.
How are law firms using AI?
The best uses save time on work a lawyer then checks.
Legal research
AI research tools answer a question in plain words and link to the cases and statutes behind the answer. A lawyer must still read every source. The tool speeds up the search, and the lawyer is still responsible for the result.
First drafts
AI can draft a demand letter, a contract clause, a client update or a section of a brief. The lawyer edits it for the facts, the law and the client's goals.
Document review and summaries
AI can read a large set of documents and pull out dates, names and key terms. It can sum up a deposition or a long contract in a page. This helps in discovery and in due diligence. My page on AI contract review covers that use in more depth.
Client intake
An AI assistant can answer the phone or the website chat, collect the basic facts and book a consult. It should never give legal advice. It should run a conflict check step before anyone sees confidential details. My guide to legal answering services explains the options.
Firm knowledge
Many firms have years of good briefs and forms spread across drives. An AI search tool can find the right precedent from your own files. My page on AI knowledge bases explains how that works.
What AI tools do law firms use?
- CoCounsel Legal: Thomson Reuters says CoCounsel combines research, analysis and drafting. It is grounded in Westlaw and Practical Law content. It also connects to Microsoft 365 and document management systems.
- Clio: Clio is practice management software for law firms. Its pricing page lists plans starting at $49 per user as of October 2026. Clio says its Manage AI add-on plans, schedules, drafts and bills for you. You ask Clio's sales team for its price.
- General AI assistants: Some firms use business versions of general AI tools for drafts and summaries. Only use plans that keep your data out of model training, and check the vendor's data terms.
For more on the software side, see my guide to law firm management software.
What does ABA Formal Opinion 512 say?
The ABA Standing Committee on Ethics and Professional Responsibility issued Formal Opinion 512 on July 29, 2024. It was the committee's first formal opinion on lawyers' use of generative AI. It says lawyers and firms should fully consider their ethical duties when they use these tools. It ties AI use to the existing Model Rules.
- Competence (Rule 1.1): A lawyer must understand the benefits and risks of the AI tools they use.
- Confidentiality (Rule 1.6): A lawyer must keep all client information confidential unless the client gives informed consent. That duty covers former and prospective clients too.
- Communication (Rule 1.4): A lawyer must reasonably consult the client about how the client's goals will be met. This can mean telling the client about AI use.
- Fees (Rule 1.5): Fees must be reasonable. A lawyer may bill for time spent entering information into an AI tool and reviewing its output. A lawyer generally cannot bill the client for time spent learning to use the tool.
Opinion 512 is guidance from the ABA. Your own state's rules of professional conduct are what bind you. Check whether your state bar has issued its own AI guidance before you set firm policy.
Does this apply differently to a solo or small firm?
The ethics duties in Opinion 512 apply the same way regardless of firm size, but the practical risk is not evenly spread. A large firm often has a dedicated knowledge management or innovation team vetting AI tools before anyone touches client work. A solo practitioner or small firm usually does not, which means the competence and confidentiality duties land directly on whichever lawyer picked the tool, often without anyone else checking the choice. If that is your situation, the single highest-value step is the one-page policy described above, written once, not the absence of a policy because the firm is small enough that it felt informal.
How do you protect client confidentiality with AI?
Treat an AI tool like any other vendor that sees client files. Before you use one, check these points:
- Data training: Make sure the vendor does not train its models on your data.
- Data storage: Find out where the data is kept and for how long.
- Access: Check who at the vendor can see your data, and whether it is encrypted.
- Client consent: Decide when you need the client's informed consent before you enter their information.
- Firm policy: Write down which tools staff may use and what they may never enter.
Free personal AI accounts are the biggest risk. Staff may paste a client email into one without thinking. A clear policy and an approved business tool cut that risk.
What are courts doing about AI-written filings?
Courts are sanctioning lawyers who file briefs with fake cases that AI made up. A New York State Bar Association article from April 2026 lists recent examples:
- Deutsche Bank National Trust Co. v. LeTennier (New York Appellate Division, Third Department, 2026): Filings across the appeal cited more than 20 non-existent, AI-generated cases, and the misuse continued even after the other side flagged it. The court imposed a $7,500 sanction on defense counsel and a further $2,500 on the client. It was New York's first appellate-level ruling on sanctions for generative AI misuse in litigation.
- Noland v. Land of the Free (California Court of Appeal, 2025): A lawyer filed 21 fake case citations and quotes. The court imposed a $10,000 sanction.
- Kruse v. Karlen (Missouri Court of Appeals, 2024): A filing had 22 fake case citations. The sanction was $10,000.
Many courts now have standing orders on AI. The US District Court for the District of Maryland has one. It says filers must check the accuracy of every filing, especially when they rely on AI. Filings with fake citations or quotes may face Rule 11 sanctions, such as striking the filing, money penalties or dismissal of the case. The judge may also refer the lawyer to the court's disciplinary committee.
Rules differ from court to court and judge to judge. Some judges ask lawyers to disclose AI use or certify that they checked it. Check the local rules and the judge's standing orders before every filing.
How do you avoid AI mistakes in court filings?
The NYSBA article gives practical steps. I would build these into any firm's process:
- Treat AI output as unchecked work: Review it the way you would review a first-year associate's draft.
- Pull every case: Read the original source for every citation and check every quote.
- Sign off before filing: Have a named lawyer approve the final filing, and keep a record of that review.
- Fix errors fast: If you find a mistake, correct it at once and be open with the court.
How should a law firm start with AI?
- Write a short AI policy: List approved tools, banned uses and who reviews AI work.
- Pick one low-risk task: Start with internal work, such as summing up public documents or drafting internal memos.
- Check the vendor: Review the data terms and security before any client data goes in.
- Train your team: Show staff how the tool works and where it fails. That also helps meet the competence duty.
- Add client work slowly: Move to client matters once your review steps work, and tell clients where that is required.
When does a custom AI tool make sense for a law firm?
Most firms should start with the AI in their research and practice management tools. A custom tool makes sense when:
- Your systems don't connect: Intake, conflicts, matter records and billing live in separate tools, and staff re-enter data.
- You have a niche practice: Your intake questions or documents are unusual, and general tools handle them poorly.
- You must control the data: You want client files processed only in systems your firm controls.
- Seat costs add up: You pay for several AI tools per lawyer, and the total keeps growing.
How would Blur Studio approach AI for a law firm?
I would start with your intake and your document workflow, since that is where most firms lose time. I build AI tools that connect to the software you already use. That might be an intake assistant that collects facts and books consults, or a search tool for your own precedents. Each tool keeps a lawyer in charge of advice and review. The firm keeps ownership of the code and all accounts. I also maintain the tools after launch.
Questions people ask
Can lawyers ethically use AI?
Yes, according to ABA Formal Opinion 512, as long as they meet their existing duties. These include competence, confidentiality, client communication and reasonable fees. State bar rules still apply, so check your state's guidance too.
Do lawyers have to tell clients they use AI?
ABA Formal Opinion 512 ties this to the duty to consult clients about how their goals will be met. In some cases a lawyer must tell the client, and entering confidential information may need the client's informed consent. Check your state bar's guidance for your situation.
What happens if a lawyer files a brief with fake AI citations?
Courts can impose sanctions. Recent cases include a $10,000 sanction in California in 2025 and a $5,000 sanction that a New York appeals court found warranted in 2026. Some courts also refer the lawyer for discipline.
Can a lawyer bill for time saved by AI?
ABA Formal Opinion 512 says fees must be reasonable. A lawyer may bill for time spent entering information into an AI tool and reviewing its output. A lawyer generally cannot bill a client for time spent learning the tool.
What is the safest first use of AI in a law firm?
Start with internal tasks that use no confidential client data, such as summing up public documents or drafting internal templates. Add client work once you have an approved tool, a written policy and a review step.
Has a solo or small firm actually been sanctioned for this?
Yes. Several of the real cases on record, including Kruse v. Karlen in Missouri, involved individual practitioners rather than large firms, which is a reminder that this is not only a large-firm risk.
Roughly what does this cost?
As a broad reference, fixes and finishes start at around $1,500, and new builds start higher, always a fixed price agreed in writing before I start, never an hourly rate. My AI app cost guide explains what actually drives a quote up or down.
Is my firm too small for this?
No. Most of the owners I work with have no tech team at all, that is the whole point. If the smarter move is a cheap software subscription instead of a custom build, I will tell you that on the first call, even though it means I do not get paid.
Sources
- ABA issues first ethics guidance on a lawyer's use of AI tools (American Bar Association)
- Avoiding sanctions in the gen AI era: practical guardrails for lawyers (New York State Bar Association)
- District of Maryland standing order regarding GenAI (Maryland State Bar Association)
- CoCounsel Legal (Thomson Reuters)
- Clio pricing (Clio)