🪶Data Minimization

NEW

Also known as: Data Reduction Principle

The principle of collecting and processing only data necessary for specified, legitimate purposes.

DESCRIPTION

Data minimization is a core privacy principle requiring that personal data be adequate, relevant, and limited to what is necessary. It shapes requirements, dataset design, and model features across systems. Implementing minimization involves scoping attributes, using privacy-preserving techniques, and enforcing retention schedules. It reduces risk, cost, and regulatory exposure while protecting individuals.

ALTERNATIVE DEFINITIONS

General Data Protection Regulation (EU)

Personal data shall be: adequate, relevant and limited to what is necessary in relation to the purposes for which they are processed.

Authoritative legal statement of the data minimization principle (Art. 5(1)(c)).

Source

UK Information Commissioner’s Office

You should only collect personal data you actually need for your specified purposes.

Plain-language guidance emphasizing necessity and purpose limitation.

Source

RELATED TERMS

Purpose Limitation Storage Limitation Privacy by Design GDPR Data Protection by Default Data Retention Anonymization Pseudonymization

COMPETING TERMS

Data Maximization

KNOWLEDGE AREAS

Data Privacy & Compliance Data Governance AI Governance

TARGET AUDIENCE

Data Governance Manager Chief Data Officer Chief Information Officer Data Steward AI Product Manager Data Management Consultant

EXTERNAL SOURCES

• Wikipedia - Data minimization
• European Data Protection Board - Guidance clarifying controller obligations including minimization
• NIST - Privacy Framework aligning technical controls to principles

TOP QUESTIONS

Why is Data Minimization important?

It reduces privacy risk, attack surface, and compliance burden. Collecting only necessary data lowers cost and improves trust.

Category: Relevancy

How do you implement Data Minimization in practice?

Define purposes, map required fields, justify each attribute, and eliminate nonessential data. Apply defaults that collect the least data.

Category: Implementation

What is a common mistake with Data Minimization?

Collecting ‘just in case’ data without a defined purpose violates principles and increases risk without delivering value.

Category: Mistakes

How does Data Minimization affect machine learning?

It focuses features on purpose-relevant signals. Consider privacy-preserving techniques to reduce identifiers while retaining utility.

Category: AI

How is Data Minimization documented and governed?

Record purpose-to-attribute mappings, DPIAs, and retention schedules. Review regularly with data stewards and privacy officers.

Category: Governance

What controls support Data Minimization?

Privacy by default settings, selective collection, tokenization, hashing, and differential privacy help enforce the principle.

Category: Controls

Is Data Minimization only about collection?

No. It spans collection, processing, storage, and sharing. Retention limits and access restrictions also apply.

Category: Scope

Does Data Minimization conflict with analytics needs?

It can, but thoughtful scoping, aggregation, and anonymization often satisfy both privacy and analytical requirements.

Category: Opinion

How often should minimization decisions be reviewed?

Review at least annually or with major changes in purpose, processing, or regulatory context.

Category: Management