Algorithms & AI

The Setting: Business as usual in pro bono settings may involve volunteer attorneys and staff working in fields less familiar to them, sometimes requiring in-depth conversations and follow-up questions with specialists, perhaps other volunteer attorneys who have received more extensive training or perhaps lawyers from another organization.  What if AI could provide the expertise? 

Generative AI platforms work by taking in information, processing it, drawing from additional background sources, and generating feedback, all based on prompts. The OpenJustice platform is customizable. Partners prime it by feeding it with rules of law and procedure based on the jurisdiction. The goal is to combine the best of a generalist AI platform with the specialized and localized application that a good lawyer can provide.  

Research Team:
Mandy Mobley Li, Assistant Director of Research Innovations, Access to Justice Lab, Harvard Law School
Samuel Dahan, Director, Conflict Analytics Lab, Queens University
Volunteer Legal Services of Central Texas
American Gateways

The Setting: Victims of domestic violence/sexual assault (DV/SA) experience the highest incidence of ancillary civil legal needs: on average, about eighteen per person in one state [this is crucial because the estimate isn’t nationwide, only from Washington State]. Yet the standard of care for civil legal assistance in the DV/SA context lags behind support for criminal prosecution. Compounding the problem is a severe lack of legal services in high-volume matters such as eviction and small-claims suits. The general expectation, therefore, is that DV/SA survivors will navigate the court system alone.

The Questions:
Does supporting DV/SA survivors in accessing available civil legal resources improve social outcomes and/or reduce revictimization? If so, could that model be replicated in other jurisdictions? This high-need, vulnerable population is difficult to study; they also have complex needs that demand the best possible allocation of limited resources. Can service providers employ research tools that indicate what really works?

The Study:
The A2J Lab has designed an evaluation to determine the impact of a unique resources referral program that assigns a legal navigator to assess survivors’ needs holistically and match them with appropriate resources. Participants will be randomized to one of two conditions: (1) automatic referral to the program; or (2) referral to other services, including direct referral to legal aid providers without the assistance of the program. The program will use a new survey tool to follow up directly with participants.

What We Hoped to Learn:
The study was expected to provide policymakers with concrete data about whether or not the program works and thus whether it is advisable to replicate (and continue to evaluate) the model elsewhere. By using new digital tools to survey DV/SA survivors, the evaluation was also expected to generate valuable data about whether or not online and text-based survey tools are effective ways to communicate with this vulnerable population.

Research Team: 
Access to Justice Lab