Statutory Interpretation: AI vs. Annotated Code
Compare AI-generated statutory interpretation against an annotated code, focusing on current-law verification, enacted-versus-proposed changes, effective dates, jurisdictional variation, and cases interpreting the statute.
Purpose
This exercise helps students see why statutory research requires a different verification posture than general legal explanation, case summaries, or secondary-source-style AI output.
Students audit an AI-generated interpretation of a statute against an annotated code. They check whether the AI used the current statutory text, identified the right subsection and exceptions, distinguished enacted amendments from proposed bills, accounted for effective dates, and noticed relevant annotations or cases interpreting the statutory language.
The point is not to make students experts in the statute. The point is to help them see that AI can make statutory law look more settled, current, or simple than it really is, especially when the tool collapses introduced bills, proposed amendments, enacted session laws, codified statutes, and effective statutory changes into one confident statement of “the law.”
Learning Goals
By the end of this exercise, students should be able to:
- Identify whether an AI-generated statutory answer is relying on current enacted law, outdated language, proposed legislation, or a not-yet-effective amendment.
- Distinguish between introduced bills, enacted amendments, codified statutory text, and effective law.
- Use an annotated code to verify statutory text, relevant subsections, definitions, exceptions, history, and cases interpreting the statute.
- Explain what the AI output got right, overstated, omitted, or treated as settled when it was not.
- Identify a concrete next step before relying on an AI-generated statutory interpretation.
Materials
- A short statutory research question.
- A pre-generated AI answer interpreting the statute.
- Access to an annotated code through Westlaw, Lexis, Bloomberg, or another legal research platform.
- A short audit worksheet with prompts: Is the statutory text current? Did the AI rely on enacted law, proposed legislation, or outdated language? What subsection, definition, exception, or effective-date provision matters most? What do the annotations add? What cases interpreting the statute should be reviewed? What would you verify next?
- Optional instructor materials: model annotated-code comparison, answer key identifying the statutory trap, and sample partner-style update.
Setup
This exercise works best after students have been introduced to the special challenges of statutory and regulatory research.
Before beginning, students should understand that statutes are structurally different from cases. Statutory research often turns on version control, effective dates, cross-references, definitions, exceptions, amendment history, and whether a change has actually become law.
The instructor should frame the exercise around a junior associate reviewing an AI-generated answer before sending it to a supervising attorney. The student’s job is not to redo the entire research project. The student’s job is to decide whether the AI’s statutory explanation can be trusted, what the annotated code confirms or complicates, and what still needs to be checked.
Run of Show
Total, approximately 30–45 minutes
- Frame the task Briefly explain that students are auditing an AI-generated statutory answer. Their goal is to determine whether the answer accurately reflects current law and whether it properly accounts for statutory structure, amendments, effective dates, and cases interpreting the statute.
- Review the AI answer Students read the pre-generated AI answer and mark any claims that depend on statutory text, amendments, definitions, exceptions, effective dates, or cases.
- Annotated-code check Students use an annotated code to locate the statute. They check the current text, relevant subsections, definitions, exceptions, history, effective-date information, and annotations.
- Bill-versus-law check Students identify whether the AI appears to rely on current codified law, superseded language, an enacted but not-yet-effective amendment, or a bill that was introduced but not enacted.
- Comparison or critique Students identify what the AI got right, what it missed, what it overstated, and what the annotated code changed or clarified.
- Documentation Students record the statute checked, the source used, the specific statutory language or annotation that mattered, and one next verification step.
- Debrief Class discussion focuses on the trap: AI may sound current and specific while blurring the difference between proposed and enacted law, current and superseded text, or statutory text and later cases interpreting it.
- Submission or reflection Students submit a short partner-style note or worksheet response explaining whether the AI answer is reliable, partially reliable, or unreliable, and why.
Student Instructions
You are a junior associate reviewing an AI-generated answer to a statutory research question. Your task is to audit the answer against an annotated code.
First, read the AI answer and identify any claim that depends on the statute’s text, an amendment, an effective date, a definition, an exception, or a case interpreting the statute.
Then use an annotated code to check:
- Whether the statutory text is current.
- Whether the AI relied on enacted law, proposed legislation, outdated language, or a not-yet-effective amendment.
- Whether the AI identified the correct subsection, definition, exception, or cross-reference.
- Whether the annotations point to cases that clarify or limit the statutory language.
- Whether there is a next source you would need to consult before relying on the answer.
As you work, keep track of:
- What the AI said the statute does.
- What the annotated code actually shows.
- Whether the difference matters.
- What you would tell a supervising attorney about whether to rely on the AI answer.
Submit a short response identifying:
- One thing the AI got right.
- One thing the annotated code clarified, complicated, or corrected.
- One verification step you would take next.
Instructor Notes
The teaching value of this exercise comes from the difference between plausible statutory explanation and verified statutory research.
Students may initially focus on whether the AI’s answer “sounds right.” Push them toward a more precise question: what legal status does the text have? Is it proposed, enacted, codified, effective, superseded, or interpreted in a way that changes the analysis?
A central teaching point is that AI may collapse the legislative process into a single statement of “the law.” An AI answer may say that a statute “now requires” something when the cited change appears only in an introduced bill. Or it may describe an enacted amendment without noticing that the effective date has not arrived. Or it may combine old statutory language with proposed amendments and present the hybrid as current law.
Common issues to watch for:
- Students may assume that a specific statutory citation means the answer is verified.
- Students may overlook effective dates.
- Students may fail to distinguish a bill from an enacted amendment.
- Students may ignore definitions or exceptions in nearby subsections.
- Students may miss annotations that point to cases interpreting the statutory language.
- Students may treat the annotated code as the end of the research instead of a map to the next source.
Strong student work usually shows:
- Clear identification of the statutory text being checked.
- Attention to whether the law is current and effective.
- Recognition of proposed-versus-enacted changes.
- Use of annotations to identify relevant interpretive cases or history.
- Candor about uncertainty.
- A concrete next verification step.
- Professional judgment rather than acceptance or rejection of the AI answer as a whole.
Adaptation
Here are some suggestions on how to adapt this in other teaching contexts. The adaptations below have been limited to firm trainings and asynchronous suggestions, but there are other possible adaptations for workshops, seminars, and other contexts. Be creative!