Practical guides, policies, and teaching resources for Penn Carey Law faculty — organized around the work you’re doing now: planning a course, designing an assessment, handling a policy question, or exploring a new teaching idea.
What are you working on?
Last updated August 7, 2026
Maintained by Polk Wagner, Deputy Dean for Academic Affairs & Innovation
Contact pwagner@law.upenn.edu
Exams & Assessment
Assessment Resources
FAQs and presentations on exam design, AI and exams, testing accommodations, and grading. These are the questions I get asked most often.
FAQ
Exams and AI — FAQ for Faculty
The questions I hear most from faculty about AI and exams — can students use it, how do you prevent it, what about offline models, and what are students saying about cheating. Includes strategies and sample exam instruction language.
What you need to know about testing accommodations — how many students receive them (~30% of 1Ls and rising), what they actually look like, what triggers them, and how they should shape your exam design decisions.
If we were starting from scratch, how would we design exams? Data on our current practices — in-class vs. take-home split, exam lengths, format breakdown — plus frameworks for rethinking assessment.
Polk Wagner · Faculty Retreat, September 2025 · Penn Law access only
How grading works here — the established grade distributions for 1L required courses, 1L electives, and upper-level; mandatory vs. suggested curves; and the A+ policy. More practical than you'd expect.
Examplify is our exam-taking software. ITS maintains the documentation — these are the pages most relevant to faculty.
ITS
Exam Administration
ITS's main exam documentation page — how exams work in Examplify, settings (including the “secure” and “no internet” options discussed in the AI FAQ), and what you need to know when setting up your exam.
Faculty should state exam rules clearly and consistently with current Registrar and University guidance. The Exams and AI FAQ offers design strategies and sample language; it does not itself authorize AI use or override controlling exam procedures.
Pedagogy & AI
Teaching in an Age of AI
Start with the course you already teach. AI may affect how students prepare, what they submit, or what an exam or paper can show you. Change only what the review shows needs changing.
Start here
Three questions about your current course
Has AI changed what students need to learn?
Has it changed how students prepare or complete the course's major work?
Has it changed what your exam, paper, or other assessment can tell you?
If the answer is no, clarify the applicable AI rules and keep the rest of the course. If preparation or final-exam conditions have changed, use the doctrinal-course guide. If research or writing has changed, use the seminar guide. Other formats and new courses are covered below.
Primary course guide
Traditional doctrinal course
Many doctrinal courses consist chiefly of class meetings followed by a final exam worth all or most of the grade. You do not need to invent regular graded assignments merely because AI exists. Review five things:
The exam questions. Do they still test the knowledge, analysis, and judgment you care about?
The exam conditions. Check in-person or take-home delivery, internet and tool access, permitted materials, uploads, local AI tools, and the controlling exam instructions.
Student preparation. Say whether students may use AI for case briefs, outlines, practice questions, or feedback on an attempted answer. Remind them that generated study material can be wrong.
Formative feedback. Revised ABA Standard 314 requires every course in the first one-third of JD credit hours to include at least one formative assessment that allows students to evaluate their performance against the course learning outcomes, beginning with the 2027–28 academic year. The method need not be graded: a practice problem, quiz, ungraded midterm, or another feedback-producing exercise may work.
Separate rules. Rules for studying, class preparation, quizzes or problems, and the final exam may differ.
Seminars vary, but many culminate in a research paper. Review four things:
Rules by stage. Address topic selection, research, outlining, drafting, and revision rather than relying only on one general AI rule.
Sources and disclosure. State what students must verify, attribute, and disclose.
What the paper shows. Decide whether the final paper still gives you enough information about the student's research, analysis, and writing.
Useful checkpoints. If needed, add a proposal, source list, outline, conference, draft, or short revision explanation. Do not collect steps you will not use or review.
Institutional requirement (POLICY-04).Revised ABA Standard 314 assessment guidance requires the assessment in affected courses, but no single method is prescribed. The method need not be graded. Current Law School implementation guidance controls local details.
Evidence note: seminar checkpoints
Design recommendation (COURSE-02; ASSIGN-03). Staging can make feedback useful and reduce the weight placed on a final paper, but it increases faculty and student workload. Use only checkpoints you will use or review.
For legal writing, clinics and externships, simulations and skills courses, mixed assessments, or a substantially redesigned course, see the additional course guidance. Clinic, placement, confidentiality, and professional rules may impose additional limits.
Make the rules clear
A short AI rules checklist
What students may and may not use AI for.
Whether the rule differs for studying, coursework, papers, and exams.
What material may not be uploaded.
What students must verify, attribute, or disclose.
Whether every student has an approved, accessible way to complete required AI work and whether a suitable alternative is needed.
Which course, exam, clinic, placement, integrity, or professional rules also apply.
Legal writing. Protect an independent first attempt when independent analysis matters; allow supervised critique or revision when revision judgment is the goal. See the first-pass revision package.
Clinics and externships. Course guidance never overrides supervision, confidentiality, client obligations, professional duties, or placement rules. Confirm permission before using a tool with matter-related information. Check current Penn guidance.
Simulations and skills courses. AI may help create scenarios or feedback, but the student's own performance and judgment must remain visible. See teaching demos.
Courses with several forms of assessment. Use a course default and repeat the controlling rule on each quiz, paper, problem, simulation, or exam. See mixed-assessment guidance.
New or substantially redesigned courses. Start with what students should know and be able to do, then decide what the exam, paper, performance, or other work should show. See the design guide.
More assessment options
What students submit: accuracy, authority, analysis, judgment, and communication when those are relevant.
How students worked: research, verification, revision, or disclosure only when those steps matter to the course goal or compliance.
What students can explain or do: a short explanation, new application, live performance, counseling, or advocacy when the paper or project alone is insufficient.
Do not add another task merely to detect AI. Use the least burdensome additional information that serves a stated purpose.
One law-school study examined four exams in four subjects at top-30 U.S. law schools and found correlations up to 0.93 between model and faculty grades when detailed rubrics were supplied. Correlation does not establish identical judgment, fairness, authorization, or validity in other courses. AI-assisted feedback or a second-pass grading check may be useful only when faculty retain responsibility, the workflow is authorized, and student work and records are handled under current Penn data guidance; the evidence does not establish that final grading should be delegated.
The AI Project Toolkit remains the source for complete packages; each Toolkit package contains instructions, disclosure language, a rubric, class-size variants, accessibility guidance, and evidence limitations.
Research and limitations
AI-assisted legal tasks. In one study, 137 upper-level law students completed six realistic legal tasks with a legal AI tool grounded in retrieved source materials, a reasoning model, or no AI. The results measured task- and system-dependent work quality or speed, not durable learning or independent competence.
Assessment Twins. This conceptual validity framework pairs interdependent tasks that address the same outcomes through different forms of evidence. It has not been empirically validated, and paired tasks can add workload, stress, or inequity.
AI detectors. A study of 192 texts across authentic pre-GenAI EFL coursework, professional writing, current-model output, and hybrid writing found that performance varies by tool, threshold, model, genre, language, and mixed authorship. A score is not standalone proof of misconduct.
AI grading. Correlations with faculty grades do not establish identical judgment, fairness, authorization, or validity in other courses. The four-exam study described above does not establish that final grading should be delegated.
Optional Faculty Tools
AI Skills for Teaching Tasks
For faculty who want to use AI in their own course-preparation workflow, I've built a set of open-source AI skills. They are designed for compatible AI coding and chat environments. For current compatibility, access, and setup information, use the AI Resources portal. Email me if you want help getting set up.
Skill
MCQ Exam Generator
Generate multiple-choice exam questions for any law course. Grounded in psychometric research — distractor validation, cognitive taxonomy tagging, and coverage balancing. Supports course presets.
Generate essay exam questions with SOLO taxonomy layering, construct alignment to your course materials, and rubrics designed for grading. Issue spotters, policy questions, cross-doctrinal fact patterns.
All-in-one class preparation: checks your slides against readings for coverage and pacing, reviews class problems, and produces a lecture guide document. Say “prep class 8” and it does the rest.
For non-teaching skills — email drafting, document comment summaries, PDF rendering — see the AI Resources portal.
AI Tools at Penn Law
Penn Law's AI tool lineup changed for 2026–27. Who has access to what, and which tools are cleared for which kinds of data, is maintained on the AI Resources portal.
Setup, eligibility, and the rules on which tools are approved for which data classifications all live on the AI Resources portal — that's the canonical spot for AI tool policies at Penn Law.
Want help getting started, a classroom demo, or to pilot AI-assisted assignments? Let me know.
Tell Me What You're Doing
I want to hear about your teaching — new things you're trying, what's working, what isn't, experiments that flopped. All of it is useful. The best ideas I've seen come from colleagues sharing what they've done, and I'd like to collect and share more of that. Drop me a note anytime: pwagner@law.upenn.edu
Presentations & Notes
Faculty Sessions and Background Materials
Sep 2025
Exams Destabilized
Polk Wagner · Faculty Retreat · September 2025 · Penn Law access only
If we were starting from scratch, how would we design exams? Data on our current practices and frameworks for rethinking assessment.
Five demos faculty can adapt — image generation, class scripts, hypothetical drafting, Virtual TA setup, and essay grading. Maintained by the Penn Carey Law AI Project.
The nuts and bolts — teaching loads, leave, course materials, and the handbooks.
Policy
Faculty Teaching & Leave System
Dean Lee's statement of teaching and leave policy — teaching loads (2 courses + seminar), teaching relief, stacking, scholarly leave, and accrual rules. Includes several changes favorable to faculty.
Dean Sophia Z. Lee · October 2024 · Penn Law access only
Faculty Teaching & Leave System — Practice Professors
Dean Lee's teaching and leave policy for the practice professor track — covers the same ground as the standing faculty memo, adapted for the differences that matter for practice faculty.
Dean Sophia Z. Lee · October 2024 · Penn Law access only
The law school's handbook for standing faculty — scholarly leave, compensation for outside activities, parental leave, disability leave, and other institutional policies. Referenced in the Teaching & Leave System memo above.
For adjunct and visiting faculty — teaching expectations, administrative procedures, exam policies, student services, and the practical things you need to know.
How to distribute course materials — PDF via Canvas is the way to go. Also covers printing options for students (Campus Copy Center has a discounted rate), library course reserve, and sample syllabus language.
For Fall 2026, the law school covers up to two paid TAs for classes expected to enroll more than 30 students. Beyond that, you can use faculty research funds or ask the Registrar about credit arrangements. Eligibility and appointment rules can change, so confirm the current process before hiring.
Paid TAs
Hiring a Paid TA
Paid TA appointments are $25/hour for Fall 2026. If you have a student in mind, have them complete the Student Worker Form. If you need to post the position, email Career Services. Questions about the process go to Mariah Ford (mford1@law.upenn.edu).
Credit TAs
TA for Credit
Credit arrangements are separate from paid appointments. Ask the Registrar's Office at reg@law.upenn.edu about current eligibility, credit limits, and the required form.
Faculty Commons
The Faculty Commons is the law school's internal portal — handbooks, policies, forms, and administrative resources. You'll need your PennKey to log in.
Everyday Resources
The “Where Do I Find...” Links
Links to pages maintained by ITS, the Library, and the Registrar. These update on their own — I'm just collecting them in one place.
ITS
Canvas — Getting Started
Setting up your course site, navigating the interface, and the basics of Canvas. Start here if you're new or need a refresher.
Research support, course reserves, purchasing requests, and your library liaison. Underused — the librarians are very good and genuinely want to help with your courses.
The academic integrity policy — what constitutes a violation, the process, and the standards. You'll want to know this exists when setting your AI and exam policies.
These routing contacts can change. If a contact or assignment changes, check Faculty Commons for the current office directory.
Exam & Academic Affairs
Questions about exam logistics, scheduling, timing, grade submission, student accommodations implementation, and Canvas site setup.
Claire Wallace · cwallac2@law.upenn.edu
Course Materials & Faculty Support
Help assembling course packs, converting materials to PDF, and general faculty support services.
Zach Siswick · siswick@law.upenn.edu
Pedagogy, AI & Curriculum
Teaching innovation, AI integration, pedagogy conversations, course design, and curriculum questions.
Polk Wagner · pwagner@law.upenn.edu
Human Resources
Leave other than scholarly leave — parental leave, disability leave, personal leave, and other HR matters.
Angela Cabrera · angelacm@law.upenn.edu
Innovation & Resources
Teaching Innovation & Further Reading
Support for your teaching, resources from across Penn, and articles worth reading.
Teaching Innovation at Penn Law
Support for Your Teaching
Two things many faculty don't know about.
Funding
Pedagogical Innovation Fund
I administer a fund to support new ideas in teaching. If you want to try something different — create a dataset or simulation for your course, produce a video, bring in a special consultant (e.g., a wellness expert), take students on an innovative field trip, or add TA support for a new teaching approach — we can help pay for it. No formal application — just email me with what you have in mind.
The Regina Austin Innovation in Teaching Award recognizes Penn Carey Law faculty whose teaching reflects innovation and thoughtful engagement with students. It honors Professor Regina Austin, whose work included critical race theory, documentary filmmaking, and public interest law. Check current faculty announcements for eligibility and nomination details.
Penn Resources
CETLI — Center for Excellence in Teaching, Learning and Innovation
Penn's university-wide teaching center. They offer consultations, workshops, and course design support. Several of their resource pages are directly relevant to us.
CETLI
Generative AI & Teaching
CETLI's guide to AI in the classroom — Penn-specific policies, assignment design, and strategies for using (or limiting) AI in your courses.
Multi-day intensive on designing (or redesigning) a course from scratch. Especially useful if you're building a new course or rethinking an existing one.
Free, confidential, one-on-one support. They'll observe your class and give you honest feedback, or help you work through a teaching challenge. Underused by law faculty.
