Policy Practicum: AI For Legal Help

Can AI increase access to justice, by helping people resolve their legal problems in more accessible, equitable, and effective ways? What are the risks that AI poses for people seeking legal guidance, that technical and policy guardrails should mitigate? In this course, students will work on teams, each of which will have a partner organization from the justice system and an interest in using AI to improve services. Partner organizations include frontline legal aid and court groups interested in using AI to improve their ability to help people dealing with evictions, criminal justice problems, debt collection, domestic violence, and other legal problems. Using human-centered design practices, students will help their partners scope out exactly where AI and other interventions might serve both the providers and the clients, what quality benchmarks should guide any new intervention, and what datasets and other projects could jumpstart a new technology initiative. Using this design work, teams will establish important guidelines to ensure that any new AI project is centered on the needs of people, and developed with a careful eye towards ethical and legal principles. This multi-stakeholder and policy research will then turn towards creative, design-driven technology development. Student teams will build a demonstration project to determine if AI is able to accomplish the legal tasks they have identified in their design research. They will consult with subject matter experts to help evaluate the AI's performance and go through iterative development cycles to refine their intervention to better meet the quality benchmarks they've established. Student teams will present their design research and technical system demo to their partners and a broader audience, for critical discussion about the next steps for the projects. They might continue to work on the efforts after the quarter or might help their partner move towards other sustainable models that could allow them to further develop, deploy, and maintain AI-powered projects to enhance their legal services. The students' learnings about engaging in responsible AI development with public interest partners will be useful to others working on AI for community agencies, government and civic tech, and high-stakes legal services. Students will be required to complete ethical training for human subjects research, which takes approximately 2 hours through the CITI program online. The students' final report will contribute to policy and technology discussions about the principles, benchmarks, and risk typologies that can guide the ethical development of AI platforms for access to justice. Students are able, but not required, to enroll in both Fall and Winter quarters of the class. The class may be extended to Spring quarter, depending on the issues raised. Elements used in grading: Attendance, Performance, Class Participation, and Written Assignments. CONSENT APPLICATION: To apply for this course, students must complete and submit a Consent Application Form available at SLS Registrar https://registrar.law.stanford.edu/. Cross-listed with the d.school (DESIGN 809E).