contents
Asynchronous (Floating) AI Work
As a 3-credit hybrid course, students should expect an intensive asynchronous component.
As per ABA Standard 310(b), students should expect to spend roughly 10 hours and 20 minutes a week engaging with this course and its materials. This includes our two-hour synchronous time together.
In addition to preparing for simulations, asynchronous work will also involve a parallel track of AI Work in which students will be asked to work with, learn about, and create AI tools. There are ten weeks of such work. This work "floats" in that studens are to work through these assignments only on weeks when they aren't activly simulating a case, shifting their due dates accordingly. More details below.
contents
Grading
As a capstone evaluation, students will write a legislative memo briefing a hypothetical legislator on AI issues. The memo will be graded using the rubric found below, and it is expected students will draw heavily from our legislative simulations.
Simulations and AI work will be graded , pass, low-pass, fail.
A student fails a simulation when the instructor finds that their performance would have been deemed negligent within the world of the simulation. This mirrors the standard used in most legal malpractice claims, but it does NOT require there to be actual harm to the hypothetical client, and the instructor will assume a reasonable student. That is, a student could win their case and still fail if the instructor believes they would have been found negligent in-game, but their actions will be measured against a reasonable student, not a reasonable practicing attorney.
AI work will be evaluated based on AI-mediated reflections. That is, every week you'll be asked to "talk" with an AI about your work and transcripts of these conversations will be reviewed by your instructor. It's just good old journaling but with an AI twist. To pass, a student's reflections must accurately reflect their work, they must show completion of the assigned work and a good-faith attempt to engage with same. Note: absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen." Aditionally, if you would like to provide traditional journal entries instead of AI-mediated reflections, you may, but you talk to me about how best to do this.
The course grade is a weighted average of a student's memo grade and their other work. It is calculated using the following weights. Note: a pass is counted as a 100%, low-pass is counted as 50%, with failures earning 0%.
Course Grade =
Simulation Stage B * 0.1 +
Simulation Stage C * 0.1 +
Simulation Stage D * 0.1 +
Weekly Reflections * 0.1 +
Legislative Memo * 0.6
This score is translated into a traditional letter grade, with grades capped at an A.
contents
Details
Let's take some time to dig into the details.
Anatomy of Simulations (Case Work)
A number of our simulations resemble tabletop role playing games like Dungeons & Dragons with the instructor acting as the game master (GM). This format allows us to inhabit various roles within the justice system without the need to make every simulation a high-fidelity dress rehearsal. There is no expectation that students have any prior experience with tabletop role playing games. Hopefully, the game mechanics will be easily understood after exposure to a few examples. See below (and in-class).
A. Context & Assignment
The case and its context will be introduced during class time. This will include a very brief orientation involving the relevant legal standards. Each case will be assigned four attorneys, two for each side of the case. Students may approach the instructor beforehand about assignments and are encouraged to do so if they foresee scheduling conflicts (e.g., they won't be able to argue on a specific date because of a pre-scheduled event like a wedding). They are also encouraged to volunteer for cases they are interested in or that they would like to avoid (e.g., because of past personal trauma). Assignments will be made randomly after taking into consideration any student preferences (i.e., excluding those for whom it wouldn't work or favoring those with an expressed preference).
Following class, all students are asked to engage with the case materials. The four student attorneys, however, are tasked with preparing their case. All other students need to be prepared to rule on questions of law relating to the case. That is, they should be ready to play the part of a fractional judge in Stage B. Unlike the attorneys, they won't need to talk, just decide.
In the time between stage A and stage B, each set of student attorneys must finalize their own case sheet. Functionally, the sheet acts as a proposed set of jury instructions (the rows in the elements column) and a witness list (subsequent columns), along with expected testimony lined up with its source and the element it aims to bolster or undermine.
Student attorneys must have their case sheet turned into the instructor, with a copy to opposing counsel, before midnight on the Friday before class.
B. Motions (including proposed jury instructions)
In class, the instructor will use the two case sheets to run a low-fidelity tabletop simulation of motion practice, acting both as judge and game master. The end goal will be to produce a single case sheet which will serve to drive the trial simulation.
The simulation might go something like this:
JUDGE/GM: I see we have agreement on all but one of the elements. Counselor, please explain to me why I should take your instruction on element three over the other side's.
ATTORNEY 1: [Makes an argument. NOTE: Since this is a low-fidelity simulation the argument doesn't have to be polished, just cogent.]
JUDGE/GM: Attorney Two?
ATTORNEY 2: [counter argument]
JUDGE/GM: [speaking to the rest of the class] How many of you think Attorney Two had the better argument?
CLASS: [Using an online polling tool the class registers their opinions. The number of affirmative votes is the argument's difficulty class (assuming 20 fractional judges). NOTE: we might have to do some math if there aren't 20 fractional judges.]
JUDGE/GM: [Decides on a modifier between 0 and 10 based on their assessment of the relative strength of Attorney 1's argument, where 0 means they didn't give it much weight and 10 means they gave it a lot of weight] Attorney one, your modifier is X. Roll.
ATTORNEY 1: [Rolls a 20-sided dice. If their roll plus their modifier is larger than or equal to the argument's difficulty class, the Judge/GM will implement some version of their suggestion]...
This basic structure will repeat for substantive questions faced by the court. These might include oral motions for summary judgment or to exclude evidence. FWIW, our simulated jurisdiction has procedural rules loosely based on the federal rules, and the GM can always clarify rules upon request. If you're making a motion based on a particular phrasing be prepared to argue that this language is the rule. The GM will make the final call. Likewise, you can argue for the inclusion of certain truth's in-game. For example, if one of the real-world parties in your case does something in real life not noted in the case materials, you can argue that it happened in-game too. The GM might shut you down quickly or give you room to argue and put it to the class. One important note: a defendant can never fully succeed on summary judgment or a motion to dismiss. We need the case to survive to trial after all. However, parts of a case may be thrown out esp. if the fates deem them as unnecessary to the course's learning objectives.
Note: students can always ask the GM to fill in details where you feel there is unnecessary ambiguity and need some context. They will not, however, do your work for you (e.g., tell you what elements to argue for in the jury instructions). Additionally, those students with non-speaking rolls (like fractional judge) may be on the class chat during simulations to ask each other questions and share insights in real time.
C. Trial
We will run a low-fidelity trial simulation in class based on the case sheet that came out of Stage B. The instructor will serve as the JUDGE/GM and the students not arguing the case will serve as the finders of fact.
There will be no opening statements. Rather, the JUDGE/GM will move through each of the sheet's columns. The simulation might go something like this:
JUDGE/GM: We have here a witness for the plaintiff, and it says they would testify to facts A, B, and C on direct. I think that's a hard bargain. You need to roll 12 or above for that all to come in. Roll.
ATTORNEY 1: [rolls 20-sided dice]
JUDGE/GM: [provides a brief narration of the outcome based on the roll (e.g., if they rolled 12 or above they describe how all of that evidence came in, under 12 how it didn't all come in)] Attorney 2, what did you hope to accomplish on cross?
ATTORNEY 2: I want to impeach them and get them to break down on the stand while screaming, "You can't handle the truth!"
JUDGE/GM: Woah, alright, you have to roll a 20 for that…
After working through all the rows, we'll break for jury deliberation at which time everyone but the student attorneys will discuss whether or not they think the plaintiff/prosecution met their burden. To make things manageable we may break into small groups and then report back with an online poll. For our purposes, a simple majority will win the day.
In class or shortly afterward, I will inform folks of the exact shape of the appeal. It is this appeal the attorneys will argue in Stage D.
D. Appeal (oral arguments)
This simulation will be of greater fidelity than Stages B and C. It will be in front of a panel of three randomly chosen student judges, drawn from students without an active case. Each attorney will be given ten minutes to speak, during which the judges may pepper them with questions and hypotheticals. The instructor will serve as DM, helping to make things run smoothly and fielding questions about the world as they come up. Attorneys are only allowed paper notes.
Immediately following arguments we will break into a whole class discussion of who won. This discussion will be led by the student judges. They will use the class as a resource to help them decide on their vote. The attorneys may not participate in this discussion.
Note: The role of student judge is not graded. It is simply assumed that as justices on our highest court you will live up to expectations. This mirrors the arrangement of US supreme court justices who lack a binding code of ethics.
contents
Anatomy of Simulations (Legislative Drafting)
The above description of Legislative Simulations should be sufficient to help you understand what this looks like. Basically, we're going to pretend to be legislators and rewrite some bills. This section is just here for symmetry and the table of contents. Consider this an example of the attention to detail you'll need to bring to our drafting. ;)
contents
Anatomy of AI Work
Nearly every week students will be assigned to create an AI workflow using the LIT Prompts extension. Most of these exercises will draw from 50 Days of LIT Prompts. This will require students to sign up for OpenAI API access and spend a nominal amount on API costs (estimated to be <$20). If this presents a financial hardship, please reach out to the instructor as limited subsidies are available. It may also be possible for students to use a free open-weight model. These assignments will include associated media consumption including readings and asynchronous lectures. There are ten weeks of such work. This work "floats" in that studens are to work through these assignments only on weeks when they aren't activly simulating a case, shifting their due dates accordingly.
Students will make clear what work they have done as part of weekly reflections. These will be AI-mediated chats (i.e., students will engage a custom LLM tool in a guided conversation about their week's work.
In addition to learning about AI and using it to build tools, students will make use of AI to interact with case materials. In addition to independent legal research, expected of student attorneys, all students will engage with cases by reading provided materials and engaging in one of the following AI interactions:
- Distill & Question a Docket. Have an AI summarize filings and ask questions of them.
- Go Socrates on a Docket. Have an AI engage the student in a Socratic dialogue based on filings.
- Moot a Case based on a Docket. Play the part of an attorney arguing one side of the case before an AI-simulated judge.
Student attorneys are expected to share their Moot interaction(s).
Students will provide their instructor with links to their AI interactions, including their weekly reflections and case-related discussions by before midnight on the Friday before class.
Devil's Advocate
Just as the role of Devil's Advocate was tasked with providing arguments against the miraculous acts of would-be saints, students should forever be on the lookout for AI "hallucinations" (a term of art roughly mapping to inaccurate statements). These tools are based on math not miracles. Consequently, any student who can point out an inaccurate statement made by an AI about a substantive matter in one of their case-related AI-produced materials will find themselves imbued with a special modifier during a simulation of their choice—making it easier for them to get their way. This includes their standard bot interactions as well as draft AI case sheets provided by the instructor. To claim this bounty, you must do more than point out an error, you must inform your instructor that you wish to make use of it.
Bug Bounty
If you discover a bug in the LIT Prompts extention, any of the prompts provided as part of the 50 days of LIT Prompts, or a simulation I've authored, and you are the first to inform me, I'll give you a LIT sticker.
contents
Office Hours/Game Time
Unless otherwise stated, office hours will take place in suite 685 on Mondays between 11am and 1pm. This time is meant as a drop in community time and will likely take place in the LIT Lab space unless a conversation warrants privacy. Use this time to talk with your instructor and fellow students about your class work, or play with LEGO and talk about your plans for the future. Most weeks, we'll have a chess board, and the instructor can be convinced to bring in a battle deck of Pokémon cards (his kids love the game so he has recently learned). Students are encouraged to bring in their own board games and coordinate with other students to reach critical mass when needed. It is hoped folks will take advantage of this time for what it is meant to be, an opportunity to learn and build community.
Of course, substantive questions about class will take precedence over "having fun," but there's no rule against asking/answering questions while playing with LEGO. Additionally, if you would like to meet at some time outside of office hours, you can book a time using the link found on our Canvas home page.
contents
Legislative Memo Grading Rubric
Students must write a memo briefing hypothetical legislators or rule makers on issues of AI relevant to one of the AI bills discussed in class. The body of the memo should not exceed 15-pages double spaced. It may be no shorter than 8 pages. It should look something like a blend of these examples: short and long writeups on California's SB 53. However, unlike the short example, it should be heavily footnoted. Additionally, you do not have to address all of a bill's provisions. We can scope your focus at our one-on-one meetings. See below.
An informal author's note, no more than 5-pages long, must be attached to each memo explaining both the how and why of their methodology for drafting the memo. That is, how was the memo written? Where did you research ideas, how did you approach copy editing, etc., and why did you make the choices you did? Yes, I want to know where you used AI, but this isn't about "catching" anyone. It's about providing you with a space to explore and decide for yourself how you want to approach questions of authorship.
You will be required to meet with me twice to talk about your paper, once before October 15th to discuss your focus, and once before the close of classes to discuss your progress. The content of these meetings should be seen primarily as a opportunity for you to calibrate your performance on coherence, usefulness, and insightfulness below.
You will be requiered to maintain a Zotero library with all of your research and to produce a table of propositions stating the proposition claimed for each cite. These will be used to inform your performance on factuality below.
Memos will be graded against five equally-weighted categories. The average score across these categories will be used as the "Memo Grade" when calculating ones final grade. The first four categories apply to the memo and the last to the author's note. The following rubric walks through the criteria:
All memos are due on December 16th at 12pm. However, given our legislative simulations at the close of the semester, there is no reason they cannot be turned in earlier. That is, if one is keeping up with class work, one will have done most, if not all, of the work needed to write their memo by the close of classes.
contents
Use of Artificial Intelligence
You should know that I believe the most immediate threat "AI" poses to education is the corrosive effect "AI policing" has on the student-teacher relationship. You should not interpret my relatively permissive AI policy to be an endorsement of the tech. It is rather an acknowledgment of the unique time we find ourselves in, one where societal norms and expectations are still in flux. My hot AI take, and a thesis of this course can be described as follows: If you are on the AI Hype Train™ I will try hard to derail you—pointing out the very real dangers people are facing in the here & now from "AI" tools. But if you dismiss the pro-social uses of "AI" out of hand, I'll point out where it can be used to help those in need.
You will be required to use AI tools in this class, and unlike other courses, you are also allowed to upload course materials into AI tools of your choice. The Law School's Rules and Regulations require all academic submissions by students to represent the student's own independent work. We will explore what this means as part of this course. See the final row in our learning objectives above.
Experience has taught me that after this course, you will likely decide not to use AI for certain tasks. That being said, there are times you will be explicitly asked not to use AI (e.g., durring the back and forth of in-class simulations), and times you are asked to use AI (e.g., in the preperation of your simulations mooting arguments etc.). I expect only that you act in good-faith and follow each assignment's understood constraints. To address any ambiguities, I ask that you are transparent and vocal about your usage or non usage of tools (e.g., in your author note). I will do the same. For example, when commenting on your weekly assignments (which will include reviewing transcripts of conversations you had with bots) I, a human, will read your transcripts and genrate a response without the aid of "AI." However, I will use "AI" to help perform an agentic test of your final paper's citations to see if they stand for the propositions for wich they were presented, using negative results to trigger closer human scrutiny. If either you or I are unsure about the other's use of AI, the expectation is that we should ask and that we will recieve a straightforward answer.
FWIW, I wrote the text of this page (and most of its HTML) without the assistance of "AI." The language exceptions are noted in-line (e.g., "LLM summary of text"). Autocompleted code suggested by my code editor is lost to history.
contents
Accommodations
For what it's worth, I made use of accommodations when I was a law student, and it was a life saver. That being said, if you anticipate issues related to the format or requirements of this course due to a disability, you should contact the Law School's Dean of Student Office for further information and assistance, including information on disability-related accommodations. We can then plan how best to coordinate any accommodations. Additionally, regardless of your accommodations status, if you're having an issue with the nature of the course materials or expectations, or anticipate having one, please let me know, and we can work to find a solution.
contents
Schedule of Classes with Assignments and Links
The law here is constantly in flux. So expect that some of the details below may change during the semester. This page will have the most up-to-date schedule and links.
Wk0: 2026-08-17
New work due on Monday 2026-08-24
Our Canvas site is live. Before we meet for the first time:
- Read our syllabus (this page you are on now) and Two Paths for A.I. (archive copy). If the article piqued your interest, you could also read AI 2027 and AI as Normal Technology (both are optional);
- Set aside 20 min and work through this doc review exercise. On completion, it will provide you with a summary file to turn in as part of the assignment linked below. Be sure to explore the Summary, Authority Details, and Full Excerpt cards by clicking on them. You can assume "(at ___)" points to the language available under Full Excerpt and that the Authority Details are correct. The assistance's determination as to whether a cite is supported is part of what we're evaluating. Additionally, your pacing buddy is not based on synchronous peer performance. It may be based on prior peer performance. If you need motivation, I will give our best preforming student a small prize from Dunkin.
- Work through this quick assignment where you can tell me a bit about yourself. It will also let you upload the results from your doc review.
turn in your assignments | contents
Wk1: 2026-08-24
New work due on Friday 2026-08-28
In class we will explore the simulation Will they show up for court?
Case Work. In preperation for next week, consider:
Bartz v. Anthropic. LLM Summary of the case:
This case is a class action lawsuit brought by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, along with their respective companies, against Anthropic PBC, the developer of the Claude AI models. The plaintiffs allege that Anthropic illegally copied and used their copyrighted books—along with hundreds of thousands of others—by training its AI systems on pirated datasets such as The Pile, which includes the Books3 collection sourced from the illegal site Bibliotik. They argue that Anthropic’s actions violated the U.S. Copyright Act by reproducing and profiting from their work without permission or compensation, leading to market harm, loss of income, and the proliferation of AI-generated content that competes with their original works. The lawsuit seeks damages and accountability for what the plaintiffs describe as a deliberate and large-scale act of copyright infringement.
Our simulation will start with facts based on the case as it stood at the filing of these documents:
FWIW, you can find the real-world docket here.
Additionally, this AI-draft case sheet is provided to help student attorneys finalize their own. It should be viewed with suspicion. Remember, "All models are wrong, but some are useful," meaning their outputs should start, not end, discussion.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk2: 2026-08-31
(no class next Monday)
New work due on Friday 2026-09-11
In class we will:
Case Work. In preperation for next week, consider:
Reddit, Inc. v. Perplexity AI. LLM Summary of the case:
Reddit is suing SerpApi, Oxylabs, AWMProxy, and Perplexity, alleging they circumvent technological protections to obtain Reddit content. Reddit claims the scraping defendants evade Reddit’s and Google’s anti-scraping measures, including by disguising automated requests, and alleges Perplexity obtains Reddit material through scraped Google search-result pages despite Reddit’s objections. Reddit seeks relief under the DMCA and related state-law theories.
Our simulation will start with facts based on the case as it stood at the filing of these documents:
FWIW, you can find the real-world docket here. Additionally, this AI-draft case sheet is provided to help student attorneys finalize their own. It should be viewed with suspicion. Remember, "All models are wrong, but some are useful," meaning their outputs should start, not end, discussion.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
Read Order on Fair Use (Bartz v Anthropic)
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk3: 2026-09-14
New work due on Friday 2026-09-18
In class we will:
- hold our motions hearing for Reddit, Inc. v. Perplexity AI
- hold our trial for Bartz v. Anthropic
Case Work. In preperation for next week:
Read The Defamation Machine
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 3 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk4: 2026-09-21
New work due on Friday 2026-09-25
In class we will:
- hold our trial for Reddit, Inc. v. Perplexity AI
- hold oral arguments for Bartz v. Anthropic appeal
Case Work. In preperation for next week, consider:
Walters v. OpenAI. LLM Summary of the case:
This case involves Mark Walters suing OpenAI for defamation after a journalist, Fred Riehl, used OpenAI's ChatGPT to summarize a legal complaint. ChatGPT erroneously generated a summary falsely stating that Walters had embezzled funds from the Second Amendment Foundation. Walters claims that these statements were entirely false and damaging to his reputation. He is seeking damages for libel. OpenAI has moved to dismiss the case, arguing that the court lacks jurisdiction and that Walters has not sufficiently established the elements of a defamation claim, including publication and actual malice.
Our simulation will start with facts based on the case as it stood at the filing of these documents:
FWIW, you can find the real-world docket here. Additionally, this AI-draft case sheet is provided to help student attorneys finalize their own. It should be viewed with suspicion. Remember, "All models are wrong, but some are useful," meaning their outputs should start, not end, discussion.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 4 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk5: 2026-09-28
New work due on Friday 2026-10-02
In class we will:
- hold our motions hearing for Walters v. OpenAI
- hold oral arguments for Reddit, Inc. v. Perplexity AI
Case Work. In preperation for next week, consider:
Anthropic PBC v. U.S. Department of War. LLM Summary of the case:
Anthropic PBC sues numerous federal agencies and officials, alleging the government unlawfully retaliated after Anthropic refused to permit Claude to be used without two restrictions: lethal autonomous warfare without human oversight and mass surveillance of Americans. The complaint challenges a presidential directive ending federal use of Anthropic technology, the Secretary of War’s supply-chain-risk designation, and related agency actions. Anthropic alleges violations of the First and Fifth Amendments, the Administrative Procedure Act, 10 U.S.C. § 3252, and limits on executive authority, and seeks declaratory and injunctive relief.
Our simulation will start with facts based on the case as it stood at the filing of these documents:
FWIW, you can find the real-world docket here. Additionally, this AI-draft case sheet is provided to help student attorneys finalize their own. It should be viewed with suspicion. Remember, "All models are wrong, but some are useful," meaning their outputs should start, not end, discussion.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
Read Order on Summary Judgement (Walters v OpenAI)
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 5 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk6: 2026-10-05
(no class next Monday)
New work due on Friday 2026-10-16
In class we will:
- hold our motions hearing for Anthropic PBC v. U.S. Department of War
- hold trial for Walters v. OpenAI
Case Work. In preperation for next week:
Read Order on Motion for Summary Judgment (Anthropic PBC v. U.S. Department of War)
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 6 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk7: 2026-10-19
New work due on Friday 2026-10-23
In class we will:
- hold our trial for Anthropic PBC v. U.S. Department of War
- hold oral arguments for Walters v. OpenAI
Case Work. In preperation for next week, consider:
Garcia v. Character Technologies. LLM Summary of the case:
Megan Garcia, on behalf of herself and her deceased son S.R.S. III, brings a wrongful death suit against Character Technologies (Character.AI), its founders, and Google. The complaint alleges that the AI chatbot service engaged S.R.S. in harmful conversations, including roleplay involving sexualized content and suicidality, ultimately contributing to his suicide at age 14. Plaintiffs assert negligence, strict liability, deceptive trade practices, and other claims, seeking damages and sweeping injunctive relief to restrict how the service operates.
Our simulation will start with facts based on the case as it stood at the filing of these documents:
FWIW, you can find the real-world docket here.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 7 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk8: 2026-10-26
New work due on Friday 2026-10-30
In class we will:
- hold our motions hearing for Garcia v. Character Technologies
- hold oral arguments for Anthropic PBC v. U.S. Department of War
Case Work. In preperation for next week, consider:
Benavides v. Tesla. LLM Summary of the case:
Neima Benavides, as personal representative of the estate of Naibel Benavides Leon, filed a wrongful death lawsuit against Tesla, Inc. in Florida federal court. The suit stems from a fatal two-car crash involving a 2019 Tesla Model S equipped with Autopilot. The plaintiff alleges that Tesla is strictly liable for manufacturing a defective and unreasonably dangerous vehicle and for failing to provide adequate warnings. Tesla removed the case to federal court citing diversity jurisdiction and denies all liability.
Our simulation will start with facts based on the case as it stood at the filing of the Answer, subsequent documents provided for inspiration:
FWIW, you can find the real-world docket here.
Download each of the above files, and use one (or more) of the tools below to engage with them. If you choose Socrates or Moot, you'll want to read through them first, and if you use Distill & Question, you'll need to read them afterward. Every student has to turn in a transcript for at least one interaction. If you've been assigned as an attorney to this case, you must turn in a transcript of your MOOT interaction(s).
Read Order on Motion to Dismiss (Garcia v Character Technologies)
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 8 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk9: 2026-11-02
New work due on Friday 2026-11-06
In class we will:
- hold our motions hearing for Benavides v. Tesla
- hold our trial for Garcia v. Character Technologies
Case Work. In preperation for next week, consider:
Legislation: In two (2) weeks our legislative session will consider the following bill: S.2630. An Act promoting economic development with emerging artificial intelligence models and safety.
Download the text of the bill, and use the tool below to engage with it.
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 9 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk10: 2026-11-09
New work due on Friday 2026-11-13
In class we will:
- hold our trial for Benavides v. Tesla
- hold oral arguments for Garcia v. Character Technologies
Case Work. In preperation for next week, consider:
Legislation: As noted last week, next week our legislative session will consider the following: S.2630. An Act promoting economic development with emerging artificial intelligence models and safety.
- Research the news coverage around S.2630, and come to class next week ready to argue: (1) for the bill as written; (2) against the bill in whole; or (3) for an amendment of your own.
"Floating" AI Work: Remember, if you are activly working a case, you don't have to do this work until after you finish your case, hence the "floating" modifier.
- Work through Week 10 of 50 Days of AI. You don't need to read anything linked to from individual posts, just read each day's post and make your own version of their Prompt Patterns. Any readings above an beyond the LIT Prompt posts will be made explicitly below.
- TBD, check back later.
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk11: 2026-11-16
New work due on Friday 2026-11-20
In class we will:
- hold oral arguments for Benavides v. Tesla
Case Work. In preperation for next week, consider:
Legislation: Come prepared to debate S.2630. Our legislative session will soon consider the following bill: H.97. An Act protecting consumers in interactions with artificial intelligence systems.
Download the text of the bill, and use the tool below to engage with it.
- Distill & Question Text (WebUI|OpenAI)
- Research the news coverage around H.97, and come to class next week ready to argue: (1) for the bill as written; (2) against the bill in whole; or (3) for an amendment of your own.
"Floating" AI Work: Work on any work that is still "floating."
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk12: 2026-11-23
New work due on SUNDAY 2026-11-28
Legislation: Come prepared to debate all of our bills.
"Floating" AI Work: Work on any work that is still "floating."
Weekly Reflection: This is how we know what you worked on. Remember, absence of evidence will be taken as evidence of absence. "If you don't mention it, it didn't happen."
turn in your assignments | contents
Wk13: 2026-11-30
Evaluation due ASAP, memo due on Wednesday 2026-12-16
Course Evaluation: Complete the evaluation available in Canvas.
"Floating" AI Work: Work on any work that is still "floating."
Legistlative Memo: Work on your memo. It is due on December 16th at noon.
complete your evaluation & turn in your final assignment/memo | contents