This course will explore connections between law and computing, with a focus on artificial intelligence and machine learning.
This is a jointly taught course, and the LAW 7127 and CS 4501 courses will meet together and mostly have the same assignments. For many of the assignments, including the final project, Law and CS students will be working closely together in a team. This syllabus applies only for the CS section, though; the LAW class has a different syllabus.
This class will be different from typical CS courses, and students in the class are expected to be open to learning new things in different ways in a different academic culture, as well as to sharing and explaining your computer science expertise with Law students. We also expect CS students to learn some new computing concepts, techniques, and skills in the course.
Expected Background. To enroll in cs4501 the expected background is:
Formal Prerequisite: completed CS 2100 (DSA1) or CS 2130 (CSO1).
Informal Prerequisite: Students are not required to have machine learning background, but should be comfortable enough with programming and math concepts to be able to go through a PyTorch tutorial on their own.
No previous background in law is expected, but students should be willing to read legal writings.
This course meets jointly with a Law School class, and CS students will work closely with Law students throughout the semester, but will have some different assignments and will be graded separately.
Meetings. Friday mornings, 9:00-11:30am in Slaughter 278 (at the Law School).
Teachers. The course will be jointly taught by Thomas Nachbar (Professor of Law) and David Evans (Professor of Computer Science).
TA. Fengyu Gao, Computer Science PhD student
Materials. Materials for the class will be posted on the course website or in canvas and will include both legal and computer science readings.
Students will need to have a subscription to poe.com, a platform that provides access to many AI tools as well as a way to create custom bots that we will use in class. A poe subscription costs approximately $20 per month and will be needed during the course to complete the assignments and exercises. (We will talk about this and expect students to setup poe.com accounts in the first class.)
Students are expected to bring a laptop to each class session (if you have a laptop problem, the CS department has a limited number of loaner machines available for students who need them).
Readings. There will be readings assigned for each class meeting that should be read before the class. Most readings will be required for both the CS and Law students, but in some cases readings that are required for the Law students that are optional for the CS students, as well as readings that are required for the CS students that are not required for the Law students. Students are expected to read and think deeply about the readings before class, and be prepared to discuss the readings during class.
Class Contribution. Students are expected to contribute actively and constructively during class sessions through discussion, dialog, and participation in in-class exercises. If you cannot attend a specific class, you should contact me (evans@virginia.edu) as early as possible. Because we will be assigning students to work in groups, it is essential that you let us know when you are unable to attend class so that we can adjust group assignments accordingly.
Individual Homework Assignments. There will be several individual homework assignments that may involve answer questions and doing exercises related to course material.
Group Projects. There will be three projects done in small teams that combine CS and Law students. These projects will involve a combination of computer science and legal techniques to address an issue in AI and ML. In addition to participating in their respective group project team, each student will write a short reflection on what they learned in conducting the project.
Final Project. For the last part of the semester, students will work in a small team on an open-ended project related to the goals of the course. There will be several intermediate deliverables, and a final presentation scheduled with the course instructors during the final exam period.
Grading. Grading is done independently for the CS4501 and LAW7127 courses. Final grades will be this default distribution:
| Item | Default |
|---|---|
| Class Contribution | 15% |
| Individual Homework | 10% |
| Group Projects 1-3 | 30% |
| Final Project (team) | 30% |
| Project Reflection Papers (individual) | 15% |
This weighting will be used to compute a minimum grade, but there is some flexibility in how different aspects of the course will be weighted to reflect exceptional performance on particular items. For example, Class Contribution can count for more than 15% for a student who makes consistently valuable contributions to the class, and an outstanding final project may overcome weaknesses in earlier assignments.
This course explicitly engages the use of AI tools as a part of its course of study and we believe that it is important for everyone take appropriate advantage of the remarkable capabilities of these tools while understanding enough about how they work to appreciate potential pitfalls and the risks of inappropriate AI tool use. Consequently, some portions of assignments will explicitly require the use of those tools while others may place constraints on their use and require you to document how you used then, and others will explicitly prohibit their use.
Each assignment will include a description of the acceptable uses of AI tools which we will aim to make clear and understandable. However, potential uses often fall into grey areas (some of which we will explore in this class), so it may not always be clear what is and is not appropriate. If at any time you have questions about whether you may use an AI tool, you should contact one of the instructors for clarification. Because we will be actively using these tools, it is not possible to provide a blanket rule, and you therefore will need to exercise particular care not to exceed boundaries for their use on different assignments in this class.
Office Hours Both instructors will have regular office hours, which will be posted on the course site soon. In addition to regular office hours, we’re happy to meet at other times if you can’t get your questions answered then or if those times don’t work for you. Please don’t hesitate to reach out to set up individual meetings.
You should also feel free to send questions to us via email. We reserve the right to post the question and response (minus any information that would identify who asked the question) to the entire class if doing so will be helpful to others. The same applies to questions asked in other forums such as office hours.
It is the University’s long-standing policy and practice to reasonably accommodate students so that they do not experience an adverse academic consequence when serious personal issues conflict with academic requirements. There are many valid reasons for accommodations and family obligations, personal crises, and extraordinary opportunities are all be potentially valid reasons for accommodations. Due to the nature of this class, though, it is important that you let your instructor know as far in advance as possible so we can discuss alternatives.
Below is an outline of the course with specific assignments for the upcoming material. Because this is a rapidly changing field and evolving course, the topics are expected to change during the course and we will confirm the readings the week before each class and post updated information as the course progresses.
Class 1 (29 August): Introduction (Cyberspace and the Law of the Horse, Introduction to LLMs)
Homework Assignment #1 Due Wednesday 3 Sept (9pm)
Class 2 (5 September): Machine Learning and Language Models
Homework Assignment #2 Due Wednesday 10 Sept (9pm)
Class 3 (12 September): Classification, Bias, and Discrimination
Start Project 1
Class 4 (19 September): Fairness, Causality, and Modification
Project 1 Due Wednesday 24 September (9pm)
Class 5 (26 September): Interpretable Models and Explainability
Project 2 Due Wednesday 1 October (9pm)
Class 6 (3 October): LLMs and Copyright
Project 3 Due Wednesday 8 October (9pm)
The second half of will focus on your Final Group Projects. Class and readings we will cover a series of topics in AI and ML connected to the law, with scheduling and topics to be determined based on student interests, availability of potential guests, and other factors.
Class 7 (10 October): Final Project Launch
Class 8 (17 October): Privacy
Class 9 (24 October): Guest: Serge Egelman
Class 10 (31 October): TBD
Class 11 (7 November): TBD
Class 12 (14 November): TBD and Draft Project Presentations
Class 13 (21 November): TBD and Draft Project Presentations
Note that 5 December is not a class day for the Law School, so we will not have a normal class meeting on Friday, 5 December, but may plan something for just the CS students.
Final Project Presentations: Scheduled with instructors during end of semester and exam period (8-16 December).