What is PII? Personal data explained
Key takeaways
- PII identifies someone directly or when combined with other details, so a dataset without names can still contain personal information.
- Pseudonymized or encrypted data remains personal data under GDPR if it can still identify someone, so removing names does not make data anonymous.
- A privacy risk assessment maps how information enters, is stored, and is deleted, who can access it, and what consequences people could face.
- Layered protections include multifactor authentication, limited access, encryption, software updates, phishing awareness, vetted outside providers, and set retention and disposal rules.
PII, or personally identifiable information, is information that identifies someone directly or can identify them when combined with other details. It includes obvious identifiers, such as Social Security numbers, but also information that becomes identifying through context. Understanding those connections is the starting point for recognizing personal data and deciding how to protect it. (csrc.nist.gov)
This article explains the terminology, common examples, relevant legal distinctions, and practical steps for handling personal information responsibly.
What is PII, and how does it relate to personal data?
PII is information that can distinguish or trace a person’s identity, either alone or alongside other information linked or linkable to that person. The National Institute of Standards and Technology, or NIST, uses this context-dependent approach: a record does not need to contain someone’s name to qualify. The important question is whether it can be connected to an individual. (csrc.nist.gov)
“Personal data” and “personal information” are related terms, but their legal meanings depend on the applicable framework. Under the European Union’s General Data Protection Regulation, or GDPR, personal data means information relating to an identified or identifiable living individual. California’s Consumer Privacy Act, or CCPA, defines personal information broadly enough to include information reasonably linkable to a consumer or household. (commission.europa.eu)
The practical takeaway is to avoid treating PII as a fixed checklist. A dataset without names may still contain personal information if account records, device identifiers, or other available details connect it to people. Review the relationships between records, not just individual fields. (csrc.nist.gov)
Recognizing personal information in everyday records
Personal information extends beyond identity documents. It can appear in purchase histories, employee files, website activity, and customer profiles. The following examples show why both the information itself and its surrounding context matter. (privacy.ca.gov)
| Information type | Examples | Practical implication |
|---|---|---|
| Identity and contact details | Names, personal email addresses, government identification numbers | These can identify people or connect separate records. |
| Digital identifiers | IP addresses, account identifiers, pseudonymous profiles | A record may relate to someone even without displaying their name. |
| Activity and relationship data | Purchases, browsing history, employment records | Information about someone’s actions can still be personal information. |
| Particularly sensitive information | Health information, precise geolocation, account access credentials | Legal requirements and potential consequences may call for additional safeguards. |
These categories illustrate common examples, not a universal legal classification. (csrc.nist.gov)
For example, imagine a retailer replaces customer names with account numbers in an order export. If another system connects those numbers to customers, removing the names has not necessarily made the export anonymous. Assess both files together before deciding how the export can be shared. (commission.europa.eu)
How do data privacy laws affect PII compliance?
PII compliance means meeting the requirements that apply to an organization’s handling of identifiable information; it is not one universal certification. Applicable duties depend on factors such as jurisdiction, organizational role, processing activities, and the information involved. Start by identifying the relevant laws rather than assuming every rule applies in the same way. (europa.eu)
Three frameworks illustrate these differences:
- GDPR: Applies to organizations established in the EU and, in specified circumstances, organizations outside it that offer goods or services to, or monitor, individuals in the EU. Obligations include having a lawful basis for processing and respecting applicable individual rights. (europa.eu)
- CCPA: Gives California residents rights concerning personal information held by covered businesses, including rights to know, delete, correct, and opt out of certain sales or sharing, subject to the law’s conditions and exceptions. (oag.ca.gov)
- HIPAA: Applies to covered entities and business associates. Its protections concern qualifying protected health information, not every piece of health-related information held by any business. (hhs.gov)
Use this overview as orientation, not a determination of your obligations. Before introducing a new collection practice or changing how information is shared, review the applicable requirements with qualified legal or privacy support. Security controls alone do not resolve questions about lawful collection, permitted use, or individual rights. (europa.eu)
Does removing names make information anonymous?
No. Removing names does not necessarily prevent identification. Under GDPR guidance, information that has been encrypted or pseudonymized remains personal data when it can still be used to identify someone; genuinely anonymous information must no longer identify an individual. (commission.europa.eu)
Pseudonymization reduces the connection between a record and a person without necessarily eliminating it. Replacing names with codes while keeping a separate matching file is a straightforward example. Anonymization, by contrast, aims to make information no longer attributable to an individual. (edpb.europa.eu)
Before sharing a supposedly anonymous dataset, examine whether other available information could reconnect it to people. Do not assume that removing a name column settles the issue. The distinction matters because information that remains identifiable may still carry data protection obligations. (commission.europa.eu)
Start personal data protection with a privacy risk assessment
A privacy risk assessment examines how processing information could create problems for individuals. NIST describes it as a process for analyzing and assessing privacy risks arising from data processing. Its focus is people and the consequences they may experience, not simply whether a system has technical weaknesses. (nist.gov)
Use these prompts to structure an initial review:
- Describe the activity. What information will you collect, and what purpose will it serve?
- Map the journey. Where will it enter, travel, remain stored, and eventually be deleted?
- Identify access. Which employees, contractors, and service providers can see or use it?
- Consider consequences. What could happen to someone if information were exposed, misused, or processed unexpectedly?
- Record improvements. Identify changes, assign responsibility, and revisit the assessment when the activity changes.
These prompts combine data inventory practices with NIST’s person-centered approach to privacy risk management. (ftc.gov)
As an illustrative exercise, consider a form that requests a full birth date merely to send a birthday greeting. Ask whether less information would meet that purpose. The goal is to make collection decisions deliberate, rather than allowing unnecessary fields to become permanent features of a workflow.
Put layered data security measures into practice
Translate the assessment into controls suited to the information and its use. The FTC recommends addressing physical security, electronic security, employee training, and service-provider practices together. Personal information in a filing cabinet deserves attention alongside information in a cloud application. (ftc.gov)
A practical starting checklist includes:
- Strengthen sign-ins. Use strong passwords, a password manager, and multifactor authentication where available. CISA recommends these as core protections for online accounts. (cisa.gov)
- Limit access. Give employees access only to information needed for their work, and remove access when responsibilities change or employment ends. (ftc.gov)
- Protect stored and transmitted information. Use appropriate encryption for sensitive information and physically secure paper files and portable media. (ftc.gov)
- Maintain software and awareness. Apply software updates and teach staff to recognize and report phishing attempts. Make reporting a suspicious message a clear, familiar action. (cisa.gov)
- Review outside providers. Investigate security practices before outsourcing work, and document expectations rather than relying only on assurances. (search.ftc.gov)
- Set retention and disposal rules. Specify how long records are needed and how to dispose of them securely afterward. (ftc.gov)
For your own rollout, give each action an owner and a way to verify completion. “Access reviewed” is more useful when it identifies the system, reviewer, date, and follow-up work. Treat the checklist as a working record, not a document filed away after approval.
Pair data breach prevention with response readiness
Data breach prevention cannot replace preparation for an incident. The FTC recommends moving quickly to secure operations, investigating the scope, preserving evidence, and fixing vulnerabilities. A response may require coordination across technical, legal, operational, and communications teams. (ftc.gov)
Prepare contact details and escalation responsibilities before information is exposed. During an incident, establish what happened, which information was involved, and who may be affected. Avoid destroying evidence while trying to restore systems. (ftc.gov)
Notification requirements vary with the circumstances and applicable laws. Obtain qualified advice promptly rather than assuming that every incident has the same reporting deadline or audience. Communications should accurately explain what is known and provide useful protective steps without exposing people to further risk. (ftc.gov)
Make responsible handling a repeatable habit
The central lesson is simple: information can identify someone through connections, not just names. Recognizing those connections is essential to deciding what needs protection. (csrc.nist.gov)
For your next review, choose one workflow and trace it from collection to disposal. Record unnecessary fields, unclear access, and unanswered legal questions. Turn those findings into specific tasks before expanding the review across the organization.
Related guides
Sources
- personally identifiable information - Glossary | CSRC
- Application of the GDPR - European Commission
- What is personal information? - privacy.ca.gov
- Data protection under GDPR - Your Europe
- California Consumer Privacy Act (CCPA) | State of California - Department of Justice - Office of the Attorney General
- hhs.gov: Index
- Anonymisation / pseudonymisation | European Data Protection Board
- Privacy Risk Assessment | NIST
- ftc.gov: Protecting personal information guide business
- Cybersecurity Best Practices | Cybersecurity and Infrastructure Security Agency CISA
- Protecting Personal Information: A Guide for Business | Federal Trade Commission
- ftc.gov: Data breach response guide business