> For the complete documentation index, see [llms.txt](https://docs.syntho.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.syntho.ai/configure-a-data-generation-job/configure-column-settings/ai-powered-generation/privacy-controls.md).

# Additional privacy controls

AI-generated synthetic data offers very high privacy levels. To maximize privacy levels with AI synthesis, Syntho provides a further set of privacy controls:

**1. Overfitting Prevention**

Prevents the model from memorizing specific patterns or properties of the original data, thus enhancing data confidentiality. During the training phase, Syntho minimizes overfitting by applying a so-called [sample noise noise ratio](/setup-workspaces/workspace-default-settings.md#privacy-control-defaults) that ensures synthetic data reflects general patterns rather than specific entries. The [privacy evaluation](/overview/get-started/ai-generated-synthetic-data.md#evaluation) metrics can also be used to demonstrate there is no overfitting.

[**2. Rare Category Protection**](/configure-a-data-generation-job/configure-column-settings/ai-powered-generation.md#rare-category-protection)

Protects the uniqueness of categorical data by substituting rare values. Rare categories, defined by a user-set threshold, are replaced with a placeholder (default: "\*"). This prevents overfitting on unique, infrequent categories and protects against potential identification based on rare data points.

[**3. Extreme Value Protection**](/configure-a-data-generation-job/configure-column-settings/ai-powered-generation.md#advanced-column-settings)

Removes outliers in numerical and date-time data to prevent re-identification based on extreme values. Outliers are detected and removed during the preprocessing phase, ensuring that potentially sensitive or identifiable extreme values do not appear in the synthetic data.

[**4. Random Noise Injection**](/setup-workspaces/workspace-default-settings.md)

Adds random noise to synthetic values to further enhance privacy. Random noise can be injected into generated synthetic data, introducing slight variations that enhance privacy while maintaining data utility. This noise factor (either absolute or relative) is configurable within the [**Workspace default settings**](/setup-workspaces/workspace-default-settings.md)**.**

[**5. Privacy evaluation**](/overview/get-started/ai-generated-synthetic-data.md#evaluation)

Syntho leverages open-source synthetic data evaluation libraries like SDMetrics to provide a transparent assessment of synthetic data quality and privacy. The platform includes an evaluation notebook that contains data quality and privacy metrics, allowing you to see how your synthetic data performs against industry standards for confidentiality and utility.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.syntho.ai/configure-a-data-generation-job/configure-column-settings/ai-powered-generation/privacy-controls.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.