Skip to content

AI for law firms: what can you use it for, and what do the rules say?

By Bhavesh Barot

Updated
Reading time
8 min read
Guide number
No. 033
A senior lawyer reads printed pages beside a young associate with an open laptop, a lime green desk lamp on the desk.
On this page
  1. 01What are law firms using AI for?
  2. 02What does ABA Formal Opinion 512 say?
  3. 03What has gone wrong already?
  4. 04How should a firm set up AI use safely?
  5. 05AI at intake: handle with care
  6. 06Using AI for marketing content
  7. 07Questions to ask an AI vendor
  8. 08How does AI change how clients find lawyers?
  9. 09How does Seen in Search help?
  10. 10Questions lawyers ask about AI

AI for law firms covers two different things. The first is using AI tools inside the firm: drafting, summarizing, research help, intake and marketing. The second is how AI changes the way clients find lawyers, now that many ask ChatGPT or read Google's AI answers before they call anyone. Both matter for a small or mid-size US firm, and both come with rules.

This guide covers the practical uses, what the ABA's 2024 ethics opinion on generative AI says, the risks that have already caught lawyers out, and how AI search affects getting found. It's general information, not legal ethics advice; your state bar's rules and opinions govern.

What are law firms using AI for?

TaskHow AI helpsMain risk
First draftsLetters, routine motions, contract clauses, client updatesErrors and invented content
SummariesLong documents, depositions, discovery setsMissing a key detail
Research starting pointsExplaining a legal concept, suggesting search termsFake or wrong citations
Document reviewSorting and flagging large setsOver-reliance without checking
IntakeChat or phone triage, schedulingGiving legal advice or forming relationships unintentionally
MarketingDrafting blog posts, FAQs, social postsInaccurate law, breaking advertising rules
AdminTranscribing, organizing notes, billing descriptionsConfidential data in the wrong tool

Legal-specific AI tools, built into research platforms and practice management software, often add safeguards such as citing sources from verified databases. General chatbots are more flexible but need more care.

What does ABA Formal Opinion 512 say?

In July 2024, the ABA issued Formal Opinion 512, its first formal guidance on lawyers' use of generative AI tools. It applies existing Model Rules rather than creating new ones. The main points:

  • Competence (Rule 1.1). Lawyers using AI need a reasonable understanding of its capabilities and limits, and must check its output. Uncritical reliance can fall short of competence.
  • Confidentiality (Rule 1.6). Before entering client information into an AI tool, lawyers must consider the risk that it could be disclosed or used. The opinion says informed client consent is needed before inputting confidential information into some self-learning tools.
  • Communication (Rule 1.4). In some situations, lawyers should tell clients about their use of AI, for example when it affects the representation significantly or the client asks.
  • Candor to the tribunal (Rules 3.1, 3.3). Lawyers must check AI-generated filings for accuracy, including citations.
  • Supervision (Rules 5.1, 5.3). Firms need policies and training on AI use by lawyers and staff, and must supervise outside vendors.
  • Fees (Rule 1.5). Lawyers billing hourly can't bill for time saved by AI as if the work were done manually. Fees must be reasonable, and general overhead costs of AI tools usually can't be passed to clients without agreement.

Many state bars have issued their own guidance too. Check yours.

What has gone wrong already?

The best-known case is Mata v. Avianca, in the Southern District of New York in 2023. Lawyers filed a brief citing cases that a chatbot had invented, then defended them when questioned. The court sanctioned the lawyers. Similar incidents have followed in other courts, and some judges now require lawyers to certify whether AI was used in a filing and that citations were checked.

The lesson isn't to avoid AI. It's that every citation, quote and statement of law from an AI tool must be checked against the actual source before it goes anywhere.

Get my free report

See whether AI assistants such as ChatGPT name your firm when clients ask for a lawyer in your area. One free report per business, with a business email you verify. No card.

How should a firm set up AI use safely?

  1. Write a short AI policy. Which tools are approved, for what tasks, and what can't be entered into them.
  2. Check data terms. For each tool, find out whether inputs are stored, used for training or seen by the vendor. Business and enterprise plans often handle data differently from free accounts.
  3. Keep client confidential information out of unapproved tools. Anonymize where possible.
  4. Verify everything. A named person checks every AI-assisted output before it's used, especially citations.
  5. Train staff. Short, practical sessions on the policy and on how AI tools fail.
  6. Review billing practices. Make sure time entries reflect actual work and your engagement letters address AI where needed.
  7. Revisit the policy as tools and your state's guidance change.

AI at intake: handle with care

AI chat on a firm's website can answer basic questions and book consultations at any hour. It also creates risks. A chatbot that answers legal questions may give legal advice, and visitors may share confidential information believing they're talking to the firm. Keep intake bots to scheduling and general information, show a clear notice that no attorney-client relationship is formed by the chat, avoid collecting detailed facts about a matter, and route anything substantive to a person. Check how the vendor stores chat data, and whether your state bar has guidance on chatbots. Review transcripts regularly to catch problems early.

Using AI for marketing content

Many firms use AI to draft blog posts and FAQs. That can save time, but your website is lawyer advertising under your state's rules, and AI drafts often contain confident errors about the law, outdated deadlines or generic statements that don't fit your state.

Treat AI drafts as rough notes. An attorney should check every statement of law, add state-specific detail and experience, and make sure claims meet Rule 7.1's ban on misleading communications. Google's guidance says it rewards helpful, reliable content however it's made, and its spam policies target mass-produced content made to manipulate rankings. Fewer, accurate, specific pages beat many generic ones.

Questions to ask an AI vendor

Before approving a tool, ask the vendor: Is our data used to train your models? Where is it stored, and for how long? Who at your company can see it? Can we delete it? Do you offer a business agreement with confidentiality terms? How does the tool cite its sources? Keep the answers on file with your AI policy, and ask again when the vendor changes its terms or releases a major update.

How does AI change how clients find lawyers?

Clients now ask ChatGPT "who's a good estate planning lawyer in Columbus" or see a Google AI summary that names a few firms, with links to sources such as legal directories, review sites, news and firm websites. You can't buy a place in those answers.

The groundwork that helps: clear practice-area pages for your state, complete directory profiles, consistent details everywhere and honest reviews. Our law firm SEO guide covers that, and our AI search engines guide explains how each assistant picks its sources. Check what AI assistants say about your firm by asking them your clients' questions, with your city, from a private window.

How does Seen in Search help?

A Seen in Search report shows whether AI assistants name your firm. It reads your homepage, writes buyer questions based on what the homepage says you offer, puts them to AI assistants such as ChatGPT, and scores how visible you are in the answers, in code, with fixed, published weights. You see each answer word for word, the competitors it named, the score, and a ranked list of what to fix first. Our methodology page shows how the score is worked out. Reports are private when they finish. It isn't legal ethics advice.

Questions lawyers ask about AI

Can lawyers use ChatGPT?
Yes, with care. ABA Formal Opinion 512 says lawyers can use generative AI if they understand its limits, protect client confidentiality, check its output and bill fairly. Avoid entering confidential client information into tools whose data terms you haven't reviewed, and verify every citation and statement of law.
What is ABA Formal Opinion 512?
It's the ABA's July 2024 formal ethics opinion on lawyers' use of generative AI. It applies existing Model Rules on competence, confidentiality, communication, candor, supervision and fees to AI tools. State bars may have their own guidance, which governs in each state.
Do lawyers have to tell clients they use AI?
Sometimes. Opinion 512 says it depends on the circumstances, such as when AI use significantly affects the representation, when confidential information goes into certain tools, or when the client asks. Some courts require disclosure in filings. Check your state's guidance and court rules.
Can lawyers bill clients for AI time?
Under Opinion 512, lawyers billing hourly should bill for actual time spent, including time reviewing AI output, not for time AI saved. Flat fees must still be reasonable. General overhead costs of AI tools usually can't be charged to clients without agreement.
What happened in Mata v. Avianca?
In 2023, lawyers in the Southern District of New York filed a brief citing cases invented by ChatGPT and initially stood by them when questioned. The court sanctioned the lawyers. The case is now a standard warning to check every AI-generated citation against real sources.
Is it safe to put client information into AI tools?
It depends on the tool's data terms. Some tools store inputs or use them for training; business plans often don't. Opinion 512 says lawyers must assess the risk and, for some self-learning tools, get informed client consent. Many firms ban confidential data in unapproved tools.
What AI tools do law firms use?
Firms use general chatbots for drafting and summarizing, AI features in legal research platforms, document review tools, transcription, practice management assistants and intake chat. Legal-specific tools often draw on verified legal databases. Approve tools by reviewing data terms and testing them on non-confidential work first.
Should a law firm have an AI policy?
Yes. Opinion 512 points to supervision duties under Rules 5.1 and 5.3, which call for policies and training. A short policy naming approved tools, permitted tasks, banned inputs and checking duties covers most of it. Review it as tools and guidance change.
Can AI write my law firm's website content?
It can help draft, but an attorney should check every statement of law, add state-specific detail and experience, and make sure claims meet your state's advertising rules. Generic, unchecked AI content risks errors and adds little value for readers or search.
How does AI search affect law firm marketing?
Clients increasingly ask AI assistants for lawyer recommendations or read AI summaries on Google. Those answers draw on directories, reviews, news and firm websites. Clear practice pages, complete directory profiles and honest reviews help you get named. Check by asking the assistants your clients' questions.
Get my free report

One free report per business, private unless you publish it. Use a business email you verify. No card. Questions about the report? Contact us.

Sources

Written by Bhavesh Barot. Last checked and updated .

Keep reading

All guides