The Frontier

Docket Literacy

Reading a court’s record directly: the docket sheet, the filings by number, the orders, and the dates. A docket tells you where a case actually stands, which a news story about it often does not.

Rulemaking Record

The trail a proposed rule or regulation leaves: the notice, the public comments, the committee reports, and the votes. For a federal court rule the record lives on uscourts.gov; for an agency rule it lives in the Federal Register and the docket on regulations.gov.

Primary and Secondary Sources

A primary source is the thing itself: the opinion, the statute, the order, the filing. A secondary source describes it. Every claim in your brief should rest on a primary source you read, with secondary sources used only to find it.

Tool Evaluation

Testing a legal AI tool against tasks with known answers and recording what it does, rather than repeating what its maker says it does. If you choose the build option, this is the standard your work is held to.

Development and Headline

A development is a dated event that changed the law or a party’s position; a headline is a story about one. Learning to tell them apart is the skill this activity is really about.

Explaining to a General Audience

Stating a legal or technical change so that someone outside the field understands what happened, why it matters, and what happens next, without losing accuracy. It is what a paralegal does for a client and what a brief does for the class.

Part 1: Choose Your Item

Pick one item from the watch list below, or propose a build project: a SKILL document, a custom assistant, or a prompt chain for a legal task you have not built before, tested against at least three inputs with known answers. If you choose a watch list item, confirm within the first day whether anything has happened since the date shown, because the list is a snapshot.

Watch list, September 2026: The New York Times fair use case against OpenAI and Microsoft (S.D.N.Y.); the revised Federal Rule of Evidence 707 and a deepfake authentication rule before the Advisory Committee on Evidence Rules; the NO FAKES Act (S. 1367) and the DEFIANCE Act (S. 275) in Congress; the Justice Department’s AI Litigation Task Force and its first challenge to a state AI law under Executive Order 14365; Colorado SB 26-189 before its January 1, 2027 effective date; Nippon Life v. OpenAI (N.D. Ill.) on unauthorized practice; Raine v. OpenAI (Cal. Super. Ct.) on chatbot liability; Mobley v. Workday (N.D. Cal.) after the opt-in period; the Florida Supreme Court or additional circuits adopting AI disclosure rules; the Nebraska Supreme Court’s final discipline in the first AI-related suspension.

Part 2: Research to Primary Sources

Use the Source Finder to locate the primary documents for your item, then read them yourself: the opinion or order, the bill text and its latest action, the docket, or the committee report. Record each source with its date. Where a secondary source makes a claim you cannot find in a primary source, leave the claim out. For a build project, the primary sources are the rule or statute your tool applies and your own test results.

Part 3: Write and Record

Write a one-page brief of 400 to 500 words that states what happened and when, why it matters to someone working in law, what the primary sources actually say, and one line on what to watch next with an expected date if there is one. Use the Explainer Scriptwriter to turn the brief into a script for a five-minute recorded explainer aimed at a classmate who has not followed the item, and record it. For a build project, use the Build Coach to assess your test results honestly before writing; the brief then describes what you built, how you tested it, and what it got wrong, and the recording demonstrates it.

Part 4: Post and Reply

Create a single new discussion thread with your brief, your source list with dates, and your recording. Then reply to two classmates’ briefs, and in each reply add one fact about their item that you verified in a primary source they did not cite, with the citation. Use the Fact Verifier to check your own replies before posting them.

Part 5: Share Chat Link

Include one AI chat link with a 1–2 sentence explanation of what the conversation shows and why you chose to share it.

Source Finder

Act as a law librarian. I will name a pending case, proposed rule, bill, or regulatory action involving artificial intelligence. Tell me where its primary sources live and how to reach them: the court and docket number and whether the docket is on CourtListener or PACER, the bill number and its page on Congress.gov, the rulemaking committee and the uscourts.gov page for its agenda books and reports, or the Federal Register citation. Give me search terms rather than guessing at document numbers or dates, and tell me clearly when you are not sure a document exists or when your information may be out of date. Do not summarize the development; I will read the sources myself.

A guide to the record, not a summary of it. Confining the model to locating sources and admitting uncertainty about dates keeps its stale knowledge from becoming your brief’s claims.

Explainer Scriptwriter

Act as a producer of short educational videos. I will paste a one-page brief on a legal development involving AI. Turn it into a script for a five-minute spoken explainer for a college classmate who has not followed the topic: an opening that says what happened in one sentence, a middle that explains why it matters and what the primary sources say, and a close that says what to watch next. Keep every factual claim exactly as it appears in my brief; do not add facts, examples, or cases I did not include. Mark any place where the brief is unclear so I can fix it before recording.

Restating without adding. The prompt lets the model do what it is good at, structure and pacing, while forbidding it to introduce facts, so the script stays as verified as the brief it came from.

Fact Verifier

Act as a fact-checker for a legal newsletter. I will paste a short claim I intend to post about a pending case, rule, or bill, along with the primary source I am relying on. Tell me whether the source supports the claim exactly as written, quoting the language that does, and flag any word in my claim that goes beyond the source, such as a date, a holding, or a prediction. If the source does not support the claim, say so. Use only the source I provide.

A last check against the text. Pairing every claim with its source and confining the model to that source is the same discipline the whole course has taught, applied to the two sentences you are about to post.

Build Coach

Act as a legal technologist reviewing a student’s build project. I will describe a SKILL document, custom assistant, or prompt chain I built for a legal task, the rule or statute it applies, and the results of testing it on at least three inputs with known answers. Help me evaluate it honestly: what it got right, what it got wrong and why, whether my test inputs were hard enough, and what one change would most improve it. Then help me describe the project in a one-page brief that states its limits as plainly as its successes. Do not praise the project; assess it.

Assessment over encouragement. A build project is only worth sharing if its failures are documented, and the prompt’s instruction to withhold praise keeps the review focused on the test results.