Home

Welcome to the ai-privacy-toolkit wiki!

The ai-privacy-toolkit contains tools and techniques related to the privacy and compliance of AI models.

The first release of this toolkit contains a single module called anonymization. This module contains methods for anonymizing ML model training data, so that when a model is retrained on the anonymized data, the model itself will also be considered anonymous. This may help exempt the model from different obligations and restrictions set out in data protection regulations such as GDPR, CCPA, etc. For more information see: https://www.ibm.com/blogs/research/2021/01/ai-privacy-boost/, https://arxiv.org/abs/2007.13086.

In the future we plan to include additional tools for applying the "right to be forgotten" to trained models without needing to retrain them from scratch, and for privacy risk assessment of ML models.

For more information about using the toolkit, see the official ai-privacy-toolkit API documentation: https://ai-privacy-toolkit.readthedocs.io/en/latest/