Synthetic Data: The Complete Series

I have spent the last year thinking and writing about synthetic data. The promise is easy to understand: generate data that looks and behaves like the real thing, without exposing anything confidential. I write from the perspective of someone working in epidemiology and healthcare research, but the principles extend beyond this area. Synthetic data is being used across finance, pharma, university research centres, and large corporate enterprises. What I have observed is that the same problems that are driving adoption across these different sectors are the same: organisations want to be able to work with more data, but the real data is sensitive in some capacity. In healthcare and finance that sensitivity is about protecting individuals from being identified, after all what could be more sensitive than your health record or bank statements. In commercial settings it is about protecting data that is valuable to competitors.

This page is the jumping off point for my synthetic data series. It is a collection of posts that I have written over the last year, covering the basics of synthetic data generation and its applications. The posts are listed in sections and represent a sensible reading order, especially if you are new to the topic. Each post stands alone and can be read independently. The series is ongoing, and I will update this page as new posts are published.

Foundations

We start with an introduction to synthetic data, some of the main methods of generation at a high level, and the applications that are driving the demand for synthetic data.

• What Is Synthetic Data and Why Does It Matter? - What synthetic data is, how it differs from anonymised or pseudonymised data, and why the distinction matters.

Methods of generation

• How Synthetic Data Gets Made - A high level overview of the different methods of generating synthetic data. Some of these methods are discussed in their own posts, and some will be in the future. Start here if you are new to synthetic data methods.

• What Are GANs and How Can They Generate Synthetic Data? - How generative adversarial networks work and why they have become one of the dominant approaches to synthetic data generation.

• Multiple Imputation and Perturbation: Why They’re Not Built for Synthetic Data - Two statistical techniques that are sometimes repurposed for synthetic data generation, and why they are not well suited to the task.

• What Are VAEs and How Can They Generate Synthetic Data - coming soon

• Diffusion Models for Synthetic Data Generation - coming soon

Applications

Here we cover the real-world use cases for synthetic data.

• How Synthetic Data Is Used in Healthcare, Research, and Beyond - Use cases across healthcare, finance, autonomous vehicles, and software testing, plus the growing institutional investment in synthetic data infrastructure.

• Why Synthetic Data is Good for Open Science - Why synthetic data has great potential for helping us do better science in a more open way.

• Synthetic Data for Rare Events - How synthetic data can be used to improve the study of rare events, and why it is important to evaluate the performance of synthetic data on smaller groups.

Evaluation

In these posts we cover some of the evaluation metrics that are important in synthetic data generation with regards to privacy, representativeness and utility.

• Representativeness in Synthetic Data: What It Means and How to Measure It - What it means for synthetic data to be representative, the four dimensions that matter, and why optimising for representativeness trades off against privacy.

• Is your Synthetic Data actually private? - How to think about privacy in synthetic data, what are the risks and what metrics can we use to measure this risk.

• How Do We Measure Utility? - Metrics and frameworks for evaluating whether a synthetic dataset is good enough for a given research question.

• The Privacy-Utility Tradeoff - How to think about the tradeoff between privacy and utility in synthetic data, and why it is important to consider both together.

Risks and failure modes

These posts are about the ways synthetic data can go wrong and what are some of the consequences of this.

• Bias in Synthetic Data - Where bias enters the generation process and why generators can amplify a skew rather than just reproduce it

• Fairness and Bias Amplification in Synthetic Data - What fairness means in synthetic data generation, how it relates to representativeness and utility, and how to detect and correct bias amplification.

• Synthetic Data in Machine Learning: Augmentation and Collapse - How synthetic data is used in machine learning, the risks of model collapse, and how to mitigate these risks.

Get in touch

This series is ongoing. If you work with synthetic data or are considering it for your research, it would be great to hear from you.

Last updated: 2 September 2026