---
title: "Data Privacy, AI, De-identification, and Anonymization: Putting It All Together"
description: In 2024, we shared several short articles about de-identified or anonymized data. Here, we take a look at how they intersect.
---

[Articles ](https://privacy-analytics.com/resources/articles)

# [Data Privacy, AI, De-identification, and Anonymization: Putting It All Together](https://privacy-analytics.com/resources/articles/data-privacy-ai-de-identification-anonymization-roundup)

 Written by [Brian Rasquinha](https://privacy-analytics.com/resources/articles/author/brian-rasquinha) | Sep 15, 2025 3:10:01 PM

In 2024, we shared several short articles about de-identified or anonymized data. Here, we take a look at how they intersect.

 

AI was a dominant topic, with much focus on the promise of AI tools and some of their inherent privacy concerns. Emerging guidance on [defensible AI](https://privacy-analytics.com/resources/articles/a-blueprint-for-defensible-ai/) includes approaches to ensure the responsible development of these tools.

 

De-identification or anonymization is a powerful privacy tool that can support defensibility in analytics, product development, and AI applications. A de-identified or anonymized dataset could be drawn from one source, but some applications require organizations to link datasets together privately without sharing key identifiers. Our [primer on tokenization and linkage concepts](https://privacy-analytics.com/resources/articles/learn-the-basics-of-data-tokenization-and-linkage/) can help you understand the challenges and solutions.

 

Even if each linked dataset is already de-identified or anonymized, the [linkage can affect the overall identifiability](https://privacy-analytics.com/resources/articles/managing-re-identification-risk-when-linking-multiple-datasets/) of the data. As such, the linkage must be managed carefully, and there are [multiple approaches](https://privacy-analytics.com/resources/articles/managing-risks-associated-with-linking-de-identified-datasets/) to consider.

 

Once you have a dataset, if you need to assess its identifiability, you might perform a re-identification risk determination (RRD), such as a HIPAA Expert Determination or similar assessment. We shared our [Best Practices for RRDs,](https://privacy-analytics.com/resources/articles/10-best-practices-for-rrds/) informed by the hundreds of such projects we perform yearly.

 

One consideration for an RRD is [the context of the data release,](https://privacy-analytics.com/resources/articles/why-context-matters-when-anonymizing-data/) which can often highly impact overall identifiability (and thus, the risks of re-identification).

 

RRDs are usually sought for structured, tabular data but can also be performed on [unstructured data like plain text,](https://privacy-analytics.com/resources/articles/work-safely-with-unstructured-text-data/) for which several organizations have built effective, defensible pipelines. With the advent of AI, there has also been an increased focus on [de-identifying DICOM medical images.](https://privacy-analytics.com/resources/articles/data-privacy-for-DICOM/)

 

One approach to managing risk as part of an RRD is to use [differential privacy techniques.](https://privacy-analytics.com/resources/articles/differential-privacy-and-risk-metrics/) These techniques inject randomness into the data to protect outputs (such as datasets or queries).

 

Finally, a de-identified or anonymized dataset can be validated for privacy protections with a [motivated intruder test.](https://privacy-analytics.com/resources/articles/motivated-intruder-testing-what-how-why/) In this test, someone tries to re-identify the data to determine whether doing so is practical.

 

Contact the experts at Privacy Analytics to learn more about data de-identification, anonymization, or other privacy topics.

[View full post](https://privacy-analytics.com/resources/articles/data-privacy-ai-de-identification-anonymization-roundup)

```json
{
  "@context" : "http://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Brian Rasquinha"
  },
  "dateModified" : "2025-09-15T15:10:01.241Z",
  "datePublished" : "2025-09-15T15:10:01Z",
  "headline" : "Data Privacy, AI, De-identification, and Anonymization: Putting It All Together",
  "image" : {
    "@type" : "ImageObject",
    "height" : 60,
    "url" : "/hs/hsstatic/content_shared_assets/static-1.4092/img/default-amp-logo.png",
    "width" : 60
  },
  "mainEntityOfPage" : "https://privacy-analytics.com/resources/articles/data-privacy-ai-de-identification-anonymization-roundup",
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "height" : 60,
      "url" : "/hs/hsstatic/content_shared_assets/static-1.4092/img/default-amp-logo.png",
      "width" : 60
    },
    "name" : "Articles"
  }
}
```