OpenMined is an open source community, with a well-documented repository on GitHub, that contributes to code-based solutions to aspects related to algorithmic safety – in particular privacy protection. The community are creating an accessible ecosystem of tools for private, secure, multi-owner governed AI by extending popular libraries like TensorFlow and PyTorch with advanced techniques in cryptography and private machine learning including: federated learning, differential privacy, multi-party computation, homomorphic encryption, consensus and threshold governance.
TensorFlow Privacy is a GitHub Library that is designed to make it easier for developers to train machine-learning models with privacy, and for researchers to advance the state of the art in MK with strong privacy guarantees. It is based on the principles of differentiated privacy. A technical white-paper is available which describes the privacy mechanisms in more detail.
