Cookie settings

Necessary, always on: they keep the site working and remember your cookie choice.

Skip to content
Skip to content

Glossary. PII and personal data.

What is PII, and how is it different from personal data?

PII is a US guidance term for information that can identify a person. This entry sets it beside personal data in the EU and UK, California, the UAE and India.

By Dipal Panchal, co-founder and CEO · Updated November 8, 2026

A US guidance term, beside five laws

Personally identifiable information (PII) is information that can distinguish or trace a person’s identity, alone or combined with other information linked or linkable to them. Unlike personal data under GDPR, which covers any information relating to an identifiable person, PII is a term from US federal guidance. The GDPR and India’s DPDP Act don’t use it, and the CCPA uses it once, by reference to a federal education-privacy law. A team that inventories only PII can miss identifiers those laws still cover.

The definition is OMB’s, in Circular A-130 (2016), and the NIST glossary repeats it; NIST SP 800-122 section 2.1 gives a longer wording for federal agencies. Each law in the table below defines its own category.

Six values under five definitions

Each value as one customer or contact record holds it, classified on November 7, 2026 against the provisions linked in the next section.

Six values under five definitions
ValueUS guidance (PII)GDPR and UK GDPRCalifornia (CCPA)UAE PDPLIndia (DPDP Act)
jordan.lee@example.comPIIPersonal dataPersonal informationPersonal dataPersonal data
Advertising ID 3f2a9c1e-…, no name attachedPII if linkable to a personPersonal data (an online identifier)Personal information (a unique identifier)Personal dataPersonal data, if the person is identifiable
Passport number (not issued)PIIPersonal dataSensitive personal informationPersonal dataPersonal data
“Type 2 diabetes” beside a customer IDPIISpecial category (health)Sensitive personal information (health)Sensitive personal data (health)Personal data; no separate sensitive class
A vendor contact’s work emailPIIPersonal dataPersonal information; the B2B exemption ended in 2023Personal dataPersonal data
“Pune subscribers: 1,240”, a countNot PIINot personal data: no one is identifiableExcluded as aggregate consumer informationNot personal dataNot personal data

Where the columns disagree

The advertising ID is the row a PII-only inventory drops. With no name attached it reads as anonymous, yet GDPR Article 4(1) names online identifiers, and Cal. Civ. Code § 1798.140 lists unique identifiers in subsection (v). The passport row splits the other way: California alone puts it in a sensitive class, under subsection (ae).

The diagnosis is sensitive under three of the five definitions. GDPR Article 9(1) and the UAE’s Decree-Law 45 of 2021, Article 1, both list health. India’s DPDP Act, section 2(t), defines personal data by identifiability and has no separate sensitive class.

India’s questions name the identifiers

At the People’s Code HR Conclave in 2026, Dipal Panchal, Sylox’s co-founder and CEO, asked HR leaders: “Where is all our Aadhaar data stored? Who has access to salary information?”

Until the DPDP Act’s main obligations commence, 18 months after the notification of November 13, 2025, the IT Act’s SPDI Rules 2011 still apply. Rule 3 lists passwords, financial information, health conditions and medical records, and biometric information as sensitive.

Use the term your law uses

In a GDPR context, saying PII invites the narrower US reading, and a team may skip data GDPR still covers, such as an online identifier. In California, say personal information; in India and the UAE, personal data.

Our rule for a global inventory: classify by data type first, such as email address, passport number or diagnosis code, then map each type to each law’s category. When a value falls under two laws, both apply, each on its own terms; our practice is to label it at the stricter level, as data classification sets out.

The label depends on the record around it

None of these labels is fixed to a value. An email address in a newsletter list and the same address in a clinic’s appointment table get the same row in this table, yet only one shows the person is a patient.

The table also doesn’t settle whether a law applies to you at all. GDPR turns on establishment and targeting, the CCPA on business thresholds, and the DPDP Act on processing digital personal data in India, or offering goods and services to people there.

With IRIS

IRIS recognizes 103 built-in data types for Indian, US and global formats. Capabilities vary by edition.

Values matched to a data type, by source, table and column: IRIS produces part of this, in the sources you connect.

Names and addresses: IRIS produces part of this, found by local name recognition, in English.

Deciding which legal category applies to a value: Not produced by IRIS.

Explore discovery and classification

Add one column per law you answer to

Take your current data inventory and add one column for each law you answer to, filled with that law’s category for each data type. The check passes when no data type is left with a blank in any column.

For the US health definition, see what is PHI. To find where each data type sits in the first place, start with the discovery step.

FAQ

What privacy teams ask about the terms

Protect your brand. Start with your data.

Bring your data inventory. We’ll show you where IRIS finds the data types it detects across the sources you connect.