Personally Identifiable Information (PII)

What is personally identifiable information (PII)?

Personally identifiable information (PII) is any information that can identify a specific person, on its own or when combined with other data. Direct identifiers include a full name, Social Security number or passport number. Indirect identifiers, such as birth date, ZIP code or job title, can identify someone when they are linked together.

For example, an HR manager uploads a spreadsheet to an AI chatbot and asks for salary trends by team. The file holds employee names, employee IDs and home addresses. The chatbot only needed the salary and team columns, but every name and address is now in the prompt. From there, the values can end up in chat history, provider logs and any system that stores the conversation.

Examples of PII

Direct identifiers Indirect identifiers
Full name Date of birth
Social Security number ZIP code or city
Passport or driver’s license number Job title and employer
Email address and phone number Gender and race
Bank account or card number IP address or device ID
Biometric data such as a fingerprint Location history

Indirect identifiers are also called quasi-identifiers. One alone rarely points to a person, but a birth date, ZIP code and gender together can narrow a large population down to a few people.

PII vs personal data vs PHI

PII is mainly a US term. The National Institute of Standards and Technology (NIST) defines it in SP 800-122 as information that can be used to distinguish or trace a person’s identity, plus information linked or linkable to that person. Other laws use different terms:

  • Personal data (GDPR): any information relating to an identified or identifiable person. It is broader than most PII definitions and covers online identifiers such as cookie IDs.
  • Personal information (CCPA): information that identifies or could reasonably be linked to a consumer or household.
  • Protected health information (PHI): identifiable health data handled by HIPAA covered entities and their business associates.

The differences matter when you decide which rules apply. This comparison of PII, PHI and PCI covers them in more detail.

Why PII matters for AI systems

PII reaches AI systems in many ways. People paste it into prompts. Retrieval-augmented generation (RAG) pulls documents that contain it. Embeddings of those documents sit in a vector database. AI agents fetch customer records through tool calls and write summaries to logs. Each step copies PII into a new place, often one that the original access controls do not cover. Finding and replacing PII before it enters these flows keeps the copies safe without stopping the work.

How Protecto helps

Protecto detects 200+ entity types across PII, PHI and PCI, plus custom entities, with detection tested in seven languages. It replaces each value with a consistent token, so models keep the context they need while real values stay protected. See PII masking for AI.

Frequently asked questions

What are examples of PII?

Common examples are names, Social Security numbers, passport numbers, email addresses, phone numbers, home addresses, dates of birth, bank account numbers and biometric data such as fingerprints.

Is an IP address PII?

Often, yes. The GDPR treats an IP address as personal data when it can be linked to a person, and US frameworks treat it as PII when it is combined with other data about someone.

What is the difference between PII and personal data?

PII is a US term focused on data that can identify a person. Personal data is the GDPR term, and it is broader, covering any information relating to an identified or identifiable person.

Related terms

Personal information · Sensitive data · Quasi-identifier · Protected health information (PHI)

Book a demo

See Protecto on your own data

Book a demo to see how Protecto finds and masks sensitive data before it reaches any AI model.

New: Protecto now runs inside LiteLLM. Control what sensitive data your models and tools see.
Read the launch blog