东盟数据匿名化指南-2025.1_53页_1mb

ASEAN Guide on Data Anonymisation Summary

Core Content

The ASEAN Guide on Data Anonymisation is a technical and application-oriented document that introduces the concept, terminology, and process of data anonymisation for personal data protection across ASEAN member states. It aims to support policymakers, regulators, and industry organisations in understanding and implementing anonymisation practices that align with their legal and operational needs.

Main Purpose

• To provide guidance on basic data anonymisation concepts and techniques.
• To serve as a baseline for adaptation to specific ASEAN jurisdictions.
• To promote harmonisation in anonymisation practices across the region.
• To support compliance with data protection laws and ensure data privacy.

Key Concepts and Terminology

Definitions

• Personal data: Data that can identify an individual, either alone or in combination with other information.
• Non-personal data: Data that does not relate to an individual.
• De-identified data: Data from which direct identifiers have been removed, nullified, or overwritten.
• Anonymisation: A risk-based process of converting personal data into non-identifiable data using techniques and governance measures.
• Anonymised data: Data that has been anonymised to achieve a low re-identification risk, meeting legal or industry standards.

Data Attribute Categorisation

• Direct identifiers (high risk): Unique attributes that can identify an individual (e.g., name, email, ID numbers).
• Indirect identifiers (medium risk): Attributes that can identify an individual when combined with other data (e.g., age, gender, postcode).
• Target attributes (low risk): Useful data that may be sensitive (e.g., financial transactions, medical diagnosis).

The Anonymisation Process

The Guide outlines a five-step process for data anonymisation:

Step 1: Know your data

• Understand the nature and use case of the data.
• Determine the suitability of data for anonymisation.
• Consider data minimisation to exclude unnecessary attributes.

Step 2: De-identify your data

• Remove direct identifiers.
• Optionally use reversible pseudonymisation to allow re-linking to the original dataset.

Step 3: Apply anonymisation techniques

• Modify indirect identifiers to prevent re-identification.
• Techniques include:
  • Character masking
  • Data generalisation
  • Data perturbation
  • Aggregation
  • Pseudonymisation
  • Attribute suppression

Step 4: Compute your risks

• Establish a risk threshold for anonymisation.
• Assess the level of re-identification risk.
• If the threshold is not met, repeat Step 3.
• Conduct a final risk assessment to determine residual risks and necessary controls.

Step 5: Manage your risks

• Apply controls and measures to further reduce re-identification risks.
• These may be contractual, administrative, or technical in nature.

Common Use Cases for Anonymisation

Benefits of Anonymisation

• Builds trust in data protection practices.
• Enables safe data sharing and collaboration without compromising privacy.
• Promotes good governance and increases consumer confidence.
• Enhances privacy and protects against data misuse.
• Reduces harm in case of data breaches (e.g., identity theft).

Scope and Limitations

• The Guide provides a general introduction to data anonymisation.
• It is not a comprehensive legal document but a technical reference.
• Focuses on tabular and structured data (e.g., Excel, SQL, JSON, CSV).
• Excludes data cleansing as a separate topic.

Data Protection Landscape in ASEAN

• Several ASEAN member states have implemented data protection laws (e.g., Singapore, Malaysia, Thailand, Philippines, Indonesia, Vietnam).
• Brunei and Cambodia are in the process of enacting similar laws.
• Anonymisation is not always legally required, but it is a best practice for data protection.
• Techniques vary by jurisdiction, but the Guide promotes a risk-based approach for harmonisation.

Common Misconceptions and Tools

• The Guide includes Annexes that address:
  • Common misunderstandings in anonymisation.
  • Advanced techniques like $ K $-anonymity, $ L $-diversity, and $ T $-closeness.
  • Anonymisation tools that can be used to implement techniques.

Conclusion

The Guide is intended to be a starting point for organisations seeking to implement data anonymisation practices in line with personal data protection principles in ASEAN. It highlights the importance of risk assessment, the categorisation of data attributes, and the application of appropriate techniques. It also underscores that anonymisation is not a one-size-fits-all process and should be tailored to specific use cases and legal requirements.