Azure Privacy Updates

[[email protected]]

With these tools, developers can ensure that their AI applications are not only secure, but also high-quality and performant.

Microsoft continues to strengthen its commitment to privacy in generative AI through several key updates to Azure . Below, we explore the key new features:

Encrypted data with the Whisper model in Azure OpenAI Service
One of the most innovative capabilities is the implementation of the Whisper model in Azure OpenAI Service, designed specifically for developers in highly regulated industries and enterprises. This functionality allows sensitive voice transcription data to be processed in the cloud , keeping the information encrypted even during processing.

This is crucial for industries that require high levels belgium telegram data of security when managing sensitive data, such as healthcare, banking and government . By keeping data encrypted at all times, you ensure that information is protected from unauthorized access and potential security breaches.

General availability of differential virtual machines (VMs) with NVIDIA H100 Tensor Core GPUs
Azure has launched general availability of confidential virtual machines (VMs) powered by NVIDIA H100 Tensor Core GPUs. These VMs enable confidential computing on GPUs, giving customers verifiable assurance that their data and code are protected throughout their lifecycle, even while being processed in CPU or GPU memory.

This capability opens up new possibilities for privacy-preserving multi-party analysis, model training, and third-party collaboration. Confidential VMs are especially useful for organizations that handle highly sensitive data and want to ensure their information is protected at all times.