3 Step Guide to Workplace Data Anonymisation

If you run a business which collects data on your employees or clients, you need to ensure that you store that data correctly. Anonymisation is a method of removing information which could reveal the identities of people within a set of data. You may need to use anonymisation as part of your data security processes and compliance with Australian privacy laws. This article outlines what it means to properly anonymise data and provides a three-step guide on how to anonymise data.

What is Anonymisation?

Anonymisation means the removal of all data features which may allow someone to identify an individual from that data. When you anonymise data, you must also consider the risk that someone could pair the data could with other information to work out the identity of someone in the data set. This risk is most likely to arise if you disclose the data to a third party.

Under Australian privacy law, data is only truely anonymised if the risk of re-identification is very low. You can make this assessment of the risk by taking into account the: 

• type of data; 
• context within which someone will use the data; and 
• circumstances where someone may disclose the data.

If you decide that anonymising data is the right choice for your business, there are three key practical steps you must take.

1. Locate All of the Identifiers

The first step in anonymising data is to review the applicable data sets you have and locate all of the information which someone could use to work out someone’s identity.

2. Choose an Anonymisation Technique

The next step is to choose an anonymisation technique. The best technique will depend on your reason for de-identifying the data and your IT capabilities. Three common techniques that you may use include: 

1. suppression;
2. generalisation; and
3. aggregation.

Suppression  

Suppression requires that you delete the identifying fields of the data. 

It is important to note that, for suppressed data, there is a risk that someone who is acquainted with a person on the list may be able to identify them based on the combination of their: 

• date of birth;
• gender;
• classification as a musician; and 
• association with your business.

This poses a risk of re-identification. To reduce this risk, you may need to also suppress the dates of birth.

Generalisation

As an alternative to suppression, you can use generalisation. Generalisation requires that you alter the identifying fields. This can reduce the re-identification risks associated with suppression, but still produce useful data. 

Aggregation

Another option is to aggregate the data. Aggregation requires that you convert the data into a summary of statistics, as illustrated in the example table below.

The aggregated data option poses the lowest risk of re-identification. However, you must also destroy the data set you used to create this aggregated data to keep the risk of re-identification low.

3. Implement Your Anonymisation Technique

Once you have isolated the identifiers and chosen your preferred anonymisation technique, the final step is to implement that technique. If you have an IT department, they may be equipped to carry this out. 

Alternatively, you may need to source external IT support to implement the anonymisation. Then, the privacy officer within your business should review the results to confirm that they have correctly executed the anonymisation.

Key Takeaways

Anonymising data can be a useful part of your business processes and assist you in meeting your privacy obligations. If you choose to anonymise data, you must: 

• locate the identifiers in the data set;
• choose a method of anonymisation; and
• properly carry this process out. 

After you have completed data anonymisation, you should check the results to confirm that the risk of re-identification is very low. If you have any questions about how to anonymise your data, contact LegalVision’s privacy lawyers on 1300 544 755 or fill out the form on this page.