AI for Law Firm Intake and Matter Setup: A Practical Guide

Why is intake a good place to start with AI?

Because it is repetitive, it touches several systems, and speed matters. A 2024 secret-shopper study in Clio's Legal Trends Report found that only 33% of the 500 law firms it emailed responded, 40% answered phone calls, and 48% were essentially unreachable by phone.¹

Much of the work behind intake is administrative. The same Clio report estimated that 81% of legal secretaries' and administrative assistants' tasks are automatable, compared to 57% of lawyers' tasks, and reported that firms that focus on improving client onboarding, for example by adding online intake tools, have 50% more incoming potential clients and earn 50% more revenue on average.¹ That is a correlation, and it doesn't show the tools caused the gain.

What does AI actually do in law firm intake and matter setup?

It handles the reading, retyping and routing, and hands each decision to a person.

Client intake is one of our most common first workflows at Soren AI. At one law firm, intake that used to take three days now takes about 20 minutes.

What do the ethics rules require for AI-assisted intake?

The same rules as manual intake, applied carefully to a faster process. Your state's rules control, so confirm the details with your ethics counsel.

• Conflicts procedures (Rules 1.7, 1.9 and 1.10): the comment to Model Rule 1.7 says a lawyer should adopt "reasonable procedures, appropriate for the size and type of firm and practice" to find conflicts, and that ignorance caused by failing to have them won't excuse a violation.² Conflicts with former clients fall under Rule 1.9,³ and Rule 1.10 generally imputes one lawyer's Rule 1.7 and 1.9 conflicts to every lawyer in the firm, with limited exceptions.⁴
• Prospective clients (Rule 1.18): a lawyer who learns information from someone who consulted about a possible representation can't use or reveal it, even if no client-lawyer relationship follows, except as Rule 1.9 would permit. If that information could be significantly harmful to the prospective client, the firm can take on an adverse client in the same or a substantially related matter only with both sides' informed consent, confirmed in writing, or if the lawyer who heard it took reasonable measures to avoid learning more than was reasonably necessary to decide whether to take the matter, is screened from it and shares no part of the fee, and the prospective client gets prompt written notice.⁵
• AI and confidentiality (ABA Formal Opinion 512): the ABA notes that Rules 1.9(c) and 1.18(b) extend similar protections to former and prospective clients' information, and says a client's informed consent is required before information relating to the representation goes into a self-learning AI tool that could disclose it.⁶

Good practice follows from these rules: collect only what is reasonably needed to check conflicts and decide whether to take the matter, keep prospective-client information in systems the firm controls, and log every conflicts decision. The rules don't prescribe these steps, so confirm the approach with your ethics counsel.

What should always stay with attorneys and staff?

Every decision with ethical or financial weight. AI can run the search across name variations and related entities, but an attorney decides whether a conflict exists, whether it can be waived and whether the firm takes the matter.

• Conflict determinations and waivers
• Whether to accept the matter
• Engagement scope, fees and terms
• Anything sent to the client in the attorney's name

The goal is for attorneys to spend their intake time on judgment calls.

Which systems does AI intake need to connect to?

The ones your intake already touches: the inbox and web forms where inquiries arrive, the practice management system that holds clients and matters, the document management system, and e-signature. Not every firm has all of these. In the ABA's 2024 Practice Management TechReport, 53% of respondents said case or practice management software is available at their firm, down from 63% in 2022.⁷

The work should fit into those systems rather than replace them, so attorneys and staff keep working where they already do. A small firm with a few inquiries a week may not need AI at all: a good intake form and a checklist in its practice management system can be enough. That is how we approach it at Soren: our engineers build into the systems a firm already uses, and we work with practice management and intake systems such as Intapp and Clio, among others. The firm owns what we build. More on AI for law firms.

How should a firm get started with AI intake?

Start by measuring and mapping what you have, then automate one step at a time.

1. Measure today's intake: time from first inquiry to response, and from response to an open matter.
2. Map the steps and systems, including who touches each step and where work waits.
3. Start with the reading and retyping: extracting details and drafting the intake summary, with people reviewing everything.
4. Add the conflicts search once the extracted data is reliable, with attorneys making every determination.
5. Automate matter setup last, after the earlier steps have run cleanly on real inquiries.

At Soren, the first two weeks go to meeting the team and interviewing the people who run intake, and by the end of week two there's a clear scope. The first workflow is typically live within two to three weeks of getting access to the systems it needs. See how we work, and for the confidentiality questions in more depth, our guide to using ChatGPT with client documents.