Finding a statute is the easy part. The hard part is knowing whether the section still says what you think it says, and whether the cases you remember interpreting it are still good law after the last amendment. AI can pull the exact text of a US Code, CFR, or state-code section in seconds, but the research win is what comes next: the amendment history and the before-and-after view of how courts read it.
This guide covers how to search statutes and regulations with AI, when to cite exactly versus search by concept, and the currency check that keeps you from quoting a section that changed two years ago.
Short answer: Search by the exact citation when you have one (42 U.S.C. section 1983) and by concept when you do not ("when is an employee eligible for FMLA"). Then check two things the text alone will not tell you: what the amendment history shows changed, and whether case law interpreting the section predates that change. A section can be live and still be interpreted by precedent that an amendment quietly retired.

TL;DR
- Exact cite first, concept second. If you have the section number, a deterministic lookup beats a semantic search that can return the adjacent subsection instead of the one you meant.
- The text is a snapshot, not the story. A statute section shows today's words. The amendment history shows what those words replaced, and when.
- Amendments come in three shapes: substitutions (old text swapped for new), insertions (new subsections added), and omissions (sections removed). Each changes what old case law is worth.
- Old precedent can outlive its statute. A case interpreting the pre-amendment text may be interpreting words that no longer exist. The before-and-after split is how you catch it.
- Coverage is US Code, CFR, and all 50 state codes. One search surface for federal statutes, federal regulations, and state law, filterable by jurisdiction, subject, and year.
Why isn't reading the current statute text enough for research?
How this was written: the capabilities below match the Statutes and Regulations feature: exact-cite lookup, semantic search, amendment history, and court interpretations split before and after each change. Every statute and amendment cited is real and checkable.
Step 1: exact cite or concept, pick the right door
There are two ways to reach a statute, and using the wrong one wastes time.
| You have | Use | Why |
|---|---|---|
| A section number ("29 C.F.R. section 825.110") | Exact lookup | Deterministic. Returns the literal section, not a near neighbor |
| A concept ("what makes an employee FMLA-eligible") | Semantic search | Understands the idea and finds the section that covers it |
| A statute name ("CCPA", "the FMLA") | Name search | Resolves the popular name to its code sections |
The reason this matters: a concept search over statutory text can return the subsection next to the one you wanted, because the words are similar. When you already know the cite, a lookup that matches the number exactly is faster and safer. AI research goes wrong most often when a semantic guess gets treated as an exact answer, so start with the cite when you have it.
Step 2: read the section, then ask what it replaced
Say you pull Cal. Civ. Code section 1798.100, a core CCPA consumer-rights provision. The current text is correct, but it is also the product of a major rewrite. The CPRA amendments, operative January 1, 2023, substituted and inserted a great deal of the California privacy statute you would have read in 2021.
That matters the moment you rely on anything written about the older version. A 2021 memo citing section 1798.100 was reading different words. The section number is the same; the law under it moved.
The amendment history is what makes that visible. Instead of one block of current text, you see the change log for the section.
| Amendment type | What it means | Research impact |
|---|---|---|
| Substitution | Old text replaced with new text | Precedent on the old wording may not carry over |
| Insertion | New subsection or proviso added | New obligations that older analysis missed entirely |
| Omission | A section or provision removed | Cases about the removed language are now hollow |
Each entry carries the amending act, its date, and the legislative reference, so you are not guessing when the change landed.
Step 3: the check that saves you, before vs after
This is the part that separates a statute browser from real research: for a given section, see how courts interpreted it before an amendment and after.
The current text is one box. The value is the split at the bottom: which precedents survived the amendment and which are interpreting words that are gone.
The court-interpretation view is deliberately strict about what counts. It surfaces opinions that reference both the section number and the act name, favors language that signals actual interpretation ("held that," "the scope of Section"), and ranks Supreme Court and circuit opinions above district-court ones. Then it splits them by whether the opinion predates or postdates each amendment.
The payoff is a question you can finally answer fast: is the case I want to cite interpreting the version of the statute that is still in force? For a fast-moving area like California non-competes, where Cal. Bus. and Prof. Code section 16600 gained new provisions in 2024 (sections 16600.1 and 16600.5), that before-and-after line is the difference between current law and a confident mistake. The California non-compete guide walks that specific change.
A worked example: the California non-compete rewrite
Here is the check proving itself on a real change. Before 2024, Cal. Bus. and Prof. Code section 16600 voided most employee non-competes, but whether California would refuse to enforce a non-compete signed in another state, under another state's law, was fought case by case. A 2022 research note might reasonably have concluded "it depends on the choice-of-law analysis."
Then the 2024 amendments landed. SB 699 inserted section 16600.5 (operative January 1, 2024), making a contract void under 16600 unenforceable no matter where or when it was signed, and barring an employer from even trying to enforce one. AB 1076 inserted section 16600.1, codifying 16600's broad reach and requiring employers to notify affected employees by February 14, 2024.
Now run the currency check on that 2022 note. Its conclusion rests on a choice-of-law balancing that section 16600.5 replaced with a flat statutory rule. The section number the note relied on is still live, so the research still looks valid, but the reasoning predates the insertion that governs the question. The before-and-after split flags the 2022 authority as pre-16600.5 and points you to the amended rule.
That is the whole point of the amendment view: not that the old note was wrong when written, but that a quiet insertion made it wrong now.
| Section | Amending act | Operative | What changed | Flag authority dated | Research consequence |
|---|---|---|---|---|---|
| Cal. Civ. Code 1798.100 (CCPA) | CPRA (Prop 24) | Jan 1, 2023 | Rights and definitions substituted and expanded | before Jan 1, 2023 | Old right-to-know analysis may read replaced text |
| Cal. Bus. and Prof. Code 16600 | SB 699 and AB 1076 | Jan 1, 2024 | Inserted 16600.5 and 16600.1: out-of-state reach, notice duty | before Jan 1, 2024 | Choice-of-law workarounds barred by statute |
| 29 C.F.R. 825.110 (FMLA) | (stable) | n/a | Eligibility test unchanged | n/a | Safe baseline; still confirm no agency update |
The FMLA row is the useful contrast. A stable section with no recent amendment means your older research on it is probably fine, and the tool showing you a clean amendment history is itself the confirmation.
Search US statutes and regulations with AI, federal and state in one place
Coverage runs across three corpora that usually live in separate tools:
- US Code: federal statutes, searchable by cite, concept, or popular name.
- Code of Federal Regulations: the agency rules that implement those statutes, where a lot of the operative detail actually sits.
- All 50 state codes: state statutes, filterable by jurisdiction, subject, and year.
For an in-house lawyer whose questions cross federal privacy law, a CFR implementing rule, and two states' versions of the same duty, one search surface beats three logins. What all-states coverage does and does not mean is covered in the 50-state research guide, and the deeper question of whether AI can do statutory interpretation, as opposed to retrieval, is in can AI interpret statutes.
What this is not
Three honest boundaries.
- It is retrieval and history, not judgment. It shows you the text, the changes, and the interpreting cases. Whether an amendment helps or hurts your position is your call, and the canons and legislative-history work stay with you.
- The public statutes API is statutes only. Vaquill AI offers a statutes API for US Code, CFR, and state codes. The court-interpretation and amendment views are product features, not an API, and there is no general case-law API.
- You still verify the cite before filing. Retrieval is a strong first step, not a substitute for the pre-filing check every brief needs. The citation verification workflow is the habit that closes the loop.
FAQ
Can AI search US Code, CFR, and state statutes together?
Yes. A single search covers federal statutes, the Code of Federal Regulations, and all 50 state codes, filterable by jurisdiction, legal subject, and year. You can search by exact citation, by concept, or by a statute's popular name.
Why does exact-cite lookup matter?
Semantic search over statutory text can return a section next to the one you meant, because the wording is similar. When you already have the section number, a deterministic lookup returns that exact section, which is faster and removes the risk of quoting the wrong subsection.
How do I know if a statute has been amended?
The amendment history for a section shows every change (substitutions, insertions, and omissions) with the amending act, its date, and the legislative reference. That change log is what tells you the current text is a rewrite of older language.
Why does an amendment affect old case law?
A case interprets the statutory text as it existed at the time. If an amendment later replaces or removes that text, the case may be interpreting words no longer in force. Splitting interpreting cases into pre-amendment and post-amendment shows which precedent still tracks the current section.
What is a court-interpretation view?
For a given statute section, it surfaces the opinions that actually interpret it: cases referencing both the section number and the act name, weighted toward interpretive language and higher courts, and split by whether they predate or postdate each amendment.
Does Vaquill AI have a case-law API?
No. The public API is statutes only (US Code, CFR, and 50-state codes). Court interpretations and amendment history are features inside the product, not an API, and there is no general case-law API.
Is AI statute search accurate enough to rely on?
Retrieval of exact text and amendment history is reliable and checkable against the source. Interpretation and the decision to rely on a given case remain yours, and every cite still goes through a verification pass before it reaches a filing.
Sources
- Cal. Civ. Code section 1798.100 (CCPA), as amended by the California Privacy Rights Act (CPRA, Proposition 24, approved November 2020), with most provisions operative January 1, 2023.
- Cal. Bus. and Prof. Code section 16600.5, added by SB 699 (2023), operative January 1, 2024.
- Cal. Bus. and Prof. Code section 16600.1, added by AB 1076 (2023), operative January 1, 2024, including the February 14, 2024 employee-notice requirement.
- 29 C.F.R. section 825.110 (FMLA eligibility), US Department of Labor regulation, used as a stable, checkable example.
New legal AI guides, weekly.
Further Reading
Legal AI Tools for Solo Practitioners: 7 Picks That Fit a One-Person Firm
Read postPersonal Injury AI in 2026: What It Actually Does for Plaintiff Lawyers
Read post7 Ways AI Helps Personal Injury Lawyers Win Cases Faster
Read postLegal API in 2026: What It Is, What It Returns, and How to Pick One
Read postAI Law Search Engine vs Keyword Search: Find Case Law Faster
Read postWhat Is NDA Triage? How Lawyers Sort Inbound NDAs in Minutes With AI
Read post
Co-Founder & CEO · Attorney
Arshita leads product and strategy at Vaquill, building the legal AI suite that solo, small-firm, and in-house US lawyers use to run a matter end to end.