AI & Law Seminar and Practicum
Course Description
6 Credits (i.e., 3 + 3)

Professor Kevin Ashley, Law 317, 412.648.1495, ashley@pitt.edu


 
 

Students in these classes will explore state-of-the-art approaches to automating legal reasoning for purposes of improving legal information retrieval, developing intelligent tutoring environments for law school students, and delivering on-line legal services. Artificial Intelligence (AI) is a branch of computer science that tries to make computers perform tasks that require intelligence if done by people, for example, reasoning and problem solving. Researchers in AI & Law design computer programs that can solve legal problems or reason about them. Both the seminar and practicum are open to law students and graduate students from other disciplines.
 

After introducing law students to some AI fundamentals and graduate students to the basics of legal reasoning, the 3-credit fall seminar focuses on reading and discussing excerpts from comprehensible papers by researchers who have built AI programs for the legal domain. The papers will sensitize law students to the techniques and assumptions employed by attorneys, and by AI and law researchers, to deal with the uncertainties inherent in legal reasoning. For instance, the papers pose tough, jurisprudential questions: Can one state precisely what a legal rule or principle means and decide whether it applies to a situation? How can one separate "hard" from "easy" legal issues? What roles do cases play in interpreting a legal rule or principle? The researchers, however, seek scientific, not philosophical, answers; they design computer programs that model legal tasks and conduct experiments to evaluate how to improve the program’s performance.
 

A seminar paper is required, which may satisfy the legal writing requirement. A partial listing of possible paper topics may be found here.  In order to stimulate discussion, students also will be asked to participate in a “rotisserie” program, a kind of software-supported peer feedback, in which the program distributes assignments to students to prepare short (1 page) critiques of readings, and then, after everyone has submitted her answer, randomly distributes these answers to different classmates for comment (and sometimes to participating authors of the readings).
 

The 3-credit practicum involves students in hands-on experiences working on AI & Law projects in a computer lab setting. The class would meet during two one-and-one-half-hour sessions each week. Depending on students’ interests and programming abilities, students would work either on their own programming projects or on some of Professor Ashley’s research projects. Students may work on projects in teams or small groups.  A Teaching Fellow and Professor Ashley will be on hand during each session (and at other times) to provide help and guidance.
 
 

Students who have some programming skills may choose to work on their own program. For instance, they may design and build a rule-based expert system for a narrow range of legal problems of the student’s choice. The student would attempt to formulate a set of expert rules for analyzing legal problems in that domain. In the process, they would learn a lot about the legal domain and would experience first hand the uncertainties inherent in legal rules, the problems of anticipating novel situations and the limitations of deduction as a model of legal reasoning. Although inevitably the resulting programs will have limitations, those limitations will likely be fundamental to a deeper understanding of legal reasoning.
 
 

Students who do not have programming skills may work on some of Ashley’s research projects. Students may, for instance, extend the case base of an on-line legal argument tutoring system to include cases in a legal domain from the first year law school curriculum or from a topic covered by the bar exam. They may help to apply an intelligent information retrieval system to a new domain of ethical or legal cases and principles. Or they may help conduct an experiment to extract information from legal texts. For these projects, no special familiarity with computer programming is required.
 
 

A Teaching Fellow or Teaching Assistant from the Graduate Program in Intelligent Systems or the Computer Science Department will assist Professor Ashley in conducting the seminar and practicum. The instructors will undertake to ensure that any programming project topics are reasonable in light of the student’s resources and likely to be educationally rewarding. The instructors will also provide students with advice and supervision in the practicum.
 
 
 

While participation in the seminar is a prerequisite for the practicum, students may take the seminar without taking the practicum. Enrollment is limited to 10 law students and 5 graduate students.
 
 

Return to Home