Legal aid programs and pro bono projects buy technology under constraints most firms never face: high volume, tight budgets, volunteers who rotate in and out, and clients whose only device may be a phone and whose evidence lives in screenshots. Tools designed for a well-staffed firm with desktop-bound clients tend to struggle in that environment, however good their feature list looks.
This guide sets priorities for program directors, managing attorneys, pro bono coordinators and the staff who run intake. It covers what to solve first, intake built for phone-only clients, safe access for volunteers, the confidentiality safeguards to require, how to evaluate cost honestly, where AI fits, and how to pilot without disrupting services.
What problems should legal aid technology solve first?
Solve the problems that consume the most staff time per client served: getting complete information and records from clients, coordinating volunteers safely, and producing materials a court can use. Reporting matters too, but good reporting depends on the first three working.
- Incomplete intake: missing facts and documents mean repeat contacts, and some clients are hard to reach a second time because of changing phone numbers, work schedules or housing instability.
- Evidence in unusable forms: screenshots, phone photos and forwarded messages that need conversion and organization before anyone can use them.
- Volunteer coordination: getting the right files to the right volunteer, and taking access away when the matter ends.
- Court-ready output: turning client records into organized exhibits for brief-service and full-representation matters alike.
- Reporting: producing the numbers funders and boards ask for without manual tallying.
Rank these for your own program before looking at any product. A housing defense clinic with same-week hearings has different priorities than a family law project with long timelines, and the right tool for one may be wrong for the other.
How do you design intake for clients who only have a phone?
Assume the client will complete intake on a phone, in short sessions, possibly on a limited data plan, with evidence that exists mainly as screenshots and photos. Every step that assumes a computer, a scanner or a long uninterrupted session will lose people.
- Requests that open from a text-message link and work in a phone browser without an app download.
- Uploads that accept screenshots and phone photos directly, including the HEIC format many iPhones use by default.
- Progress that saves, so a client can stop and return later.
- Plain-language instructions, with translations for the languages your community speaks.
- Specific requests ("photos of the notice taped to your door, front and back") rather than "send your evidence."
- Originals preserved on arrival, because many clients cannot resubmit after a phone is lost, replaced or reset.
Specific requests matter more than any single feature. CourtBriefly, for instance, lets staff request specific records from a client, name the recipient, keep submissions attached to the matter and track what is still missing, and its evidence intake tools normalize HEIC photos and hash originals on arrival. Whatever you adopt, test it with real clients on inexpensive phones before rollout. The client evidence submission portal guide covers what to look for in more depth, and the screenshot triage workflow helps when a client sends everything at once.
How should volunteer and pro bono attorneys get access?
Give each volunteer named, scoped access to only the matters they handle, for only as long as they handle them, with supervising staff kept in the loop. Forwarding files to a volunteer's personal email is the pattern to eliminate.
- Grant access per matter. Volunteers should not see the program's full caseload.
- Release selected materials, not whole files, when a volunteer needs only part of the record, and set an expiry date.
- Record acknowledgment that the volunteer received the package.
- Keep internal discussion separate from anything shared with the client.
- Revoke access at matter close as a standard closing step, not an afterthought.
Onboarding matters as much as access control, because volunteers rotate. A one-page guide covering how to open an assigned matter, where client materials live, what may and may not be downloaded, and whom to ask, will save staff from answering the same questions every clinic night. A practice matter loaded with test data lets new volunteers learn the system without touching anyone's real case.
Supervision duties still apply. Staff attorneys supervising volunteers and nonlawyer staff carry responsibilities under Model Rules 5.1 and 5.3. For brief-service clinics, Model Rule 6.5 relaxes certain conflict rules for short-term limited legal services provided through nonprofit or court-annexed programs, applying them only when the lawyer knows of a conflict; state versions vary, and the relaxed standard stops applying if the volunteer takes on continuing representation. A limited-scope representation workflow covers scope letters and handoffs.
What confidentiality safeguards should a program require?
Require safeguards proportionate to the sensitivity of the information, which in legal aid is often very high: housing, immigration, public benefits, health and family safety information. Model Rule 1.6(c) requires reasonable efforts to prevent unauthorized access, and its comment 18 weighs factors such as the sensitivity of the information, the likelihood of disclosure without added safeguards, and the cost and difficulty of those safeguards.
- Encryption in transit and at rest, and a clear answer on where data is stored.
- Role-based access, with separation by client or matter.
- Credentials that can be scoped, set to expire and revoked.
- Records of who accessed and released what, and when.
- Contract terms on breach notification, data use (including no use of client data for model training) and deletion at the end of the relationship.
- Data export in standard formats, so the program is not locked in when a grant ends.
For a broader treatment of vendor diligence, see client confidentiality and cloud software.
How should a program evaluate cost?
Evaluate total cost over the life of the tool, not just the license price. Staff time to administer, train and work around a tool can exceed the subscription, and volunteer turnover multiplies training costs.
| Cost element | Question to ask |
|---|---|
| Licensing | How does price scale with clients, matters, staff and volunteers? |
| Administration | Who sets up matters, users and access, and how long does it take each time? |
| Training | How quickly can a rotating volunteer become productive? |
| Integration | Does it connect to your case management system, or duplicate data entry? |
| Exit | Can you export everything if funding changes? |
Ask vendors for published pricing and a free trial you can run with test data, rather than relying on a sales demo. Then build a simple three-year cost estimate that includes staff hours, and compare it with the staff time the tool is expected to save. If the savings depend on everyone changing how they work at once, plan for a slower ramp than the vendor suggests.
Where does AI fit in a legal aid program?
AI can reduce the staff time spent on mechanical steps, such as reading text out of screenshots or drafting a first-pass chronology, but its output needs human review, and the confidentiality analysis comes first. ABA Formal Opinion 512 is the starting point, along with any guidance from your state bar.
The opinion concludes that informed client consent is needed before information relating to a representation is entered into a self-learning AI tool that could disclose it, and it expects supervisors to set policies and train the people using these tools. In a legal aid setting, that points toward tools that do not train on client data, clear rules on which tasks AI may assist with, and review of every AI-assisted output by a person who checks it against the source. Keep the steps that must be exact, such as exhibit numbering and document fingerprints, free of AI entirely.
How do you pilot new technology without disrupting services?
Pilot with one clinic, office or case type for a defined period, measure against a baseline, and decide on evidence. A program-wide rollout of an untested tool puts the service itself at risk.
- Pick a contained scope, such as one housing clinic or one family law intake stream.
- Measure the baseline: staff time per intake, repeat client contacts, and time to court-ready materials.
- Train the pilot group, including volunteers, with short written guides.
- Run for 60 to 90 days, collecting staff and client feedback along the way.
- Compare against the baseline and decide: expand, adjust or stop.
Before adopting any tool that changes how client information is handled, have your program's ethics or risk lead, or outside ethics counsel licensed in your state, review the configuration and the vendor terms.
