Is it legally
a data breach?
The same exposed data can be covered in one state and fall outside the listed categories in another. Explore what breach notification laws cover across 54 jurisdictions.
There is no single comprehensive national breach notification law. Federal obligations are sector-specific.
Provisional research edition · Research cutoff: October 29, 2026
Breach Simulator
Choose a breach, or pick the kinds of data exposed. The map shows where those fields fall inside each law's covered categories.
How to read the results
How would these reported fields map to the laws in our October 2026 snapshot?
A match is a starting point. Blue “requires review” results flag missing incident facts, not weaker legal coverage. Historical incidents are compared with the October 2026 law snapshot.
Build a hypothetical breach
Field by field: what matched and what didn't
Select a state for its categories, score range and sources. With no breach chosen, the map compares overall breadth or one category.
What our breach records show
We joined every breach in the ObscureIQ Breach Risk Index whose data we have verified circulating to this index, using the fields we observed in circulation rather than the fields a company reported. Below are the totals and four representative cases. A missing public notice is not proof that no notice was sent or that a law was violated.
The same stolen data can be a reportable breach in one state and nothing in the next, and much of the data that causes real harm sits outside every statute. These four cases from our records show the pattern. Each can be opened in the simulator.
Case-level tables for all 269 breaches, including those with no public notice found, are available to privacy advocates, regulators and legislators on request. Contact ObscureIQ.
Jurisdiction ranking
All 54 jurisdictions, searchable and sortable
Search and sort both tables. Select a jurisdiction for categories, recorded score ranges, qualifications and sources.
| Rank | Jurisdiction | Score | Currency review |
|---|
Category coverage matrix
Every jurisdiction against every category
1 = full model credit (green) · ½ = narrow / conditional credit (amber) · 0 = no assigned credit (red). Zero does not mean no legal protection. The last three columns are trigger features. Scroll horizontally.
How the index works
What the score measures
Each of the 54 jurisdictions (the 50 states, the District of Columbia, Puerto Rico, Guam and the US Virgin Islands) gets a score from 0 to 100 for how much personal information its breach-notification law covers. A higher score means more kinds of stolen data oblige a company to tell the people affected.
Most of the score, 80 points, comes from twelve kinds of information: Social Security numbers, driver's licenses and state IDs, other government IDs, financial account data, medical information, health insurance information, login credentials, biometric data, tax IDs and PINs, precise location, genetic data, and other identifiers. Each is worth the same. The other 20 points cover how the law treats combinations: whether a name plus one of those items counts (5 points, the same in every jurisdiction), whether a username or email with a password counts on its own (8), and whether some items count even without a name (7).
Each item earns full, half or no credit. Half credit means the law covers it narrowly or only under conditions. Jurisdictions with the same score share a rank.
What it does not measure
The score measures breadth, not overall privacy protection, and it does not decide whether a particular breach required notice. Real outcomes also turn on things the score leaves out: encryption safe harbors, exceptions where harm is unlikely, and separate federal or sector rules such as HIPAA and GLBA, unless a state folds them into its main breach law.
How current it is
The laws are as of October 29, 2026. We checked each one for later amendments: for 18 jurisdictions that check is complete within limits noted in the dataset, and for the other 36 we have not yet confirmed every later amendment. Each entry shows its status. Each jurisdiction's score range reflects nine reasonable ways of reading its law. Some source links may have moved since we checked them.
The breach simulator
Pick the kinds of data in a breach, or choose one of the example breaches, and the simulator checks each jurisdiction's categories. There are four results: a category match, a conditional match, needs review, and no match. A match means the data falls in a covered category. It does not mean notice was legally required, because the simulator does not apply the exceptions above.
The example breaches use the data we have seen circulating in breach indexes (Have I Been Pwned and DataBreach.com), not the list a company reported, because the question is what the stolen data would trigger. A plain-text password with a username or email counts as login credentials. A hashed or encrypted password needs review, and we do not assume it was cracked. Partial card numbers and partial Social Security numbers are not counted. The "names are linked" box is an assumption you set, not a finding about the data.
What "no notice found" means
It means we found no notice in state attorney-general filings, the federal HHS breach portal or company statements. Some notices go only to the people affected and are never published, so a missing notice is a lead, not proof that a company broke the law. Companies outside the US owe state notice only to their US residents.