Microsoft Injects ChatGPT into Secure US Government Cloud

Microsoft has launched its Azure OpenAI service, which enables Azure government customers to access GPT-3 and 4 alongside Embeddings. The ChatGPT programme is being offered to government agencies without compromising Microsoft‘s secure standards for meeting government data requirements. Microsoft promises that the service will securely access large language models in the commercial environment via REST APIs, a Python SDK, or Azure AI Studio without exposing government data to the public internet. Additionally, Microsoft encrypts all Azure traffic using IEEE 802.1AE, or MACsec, network security and it guarantees that all traffic will remain within its global backbone of 250,000 km fibre optic and undersea satellite cable systems. Azure OpenAI Service for government is now available to approved enterprise and government customers. However, Microsoft and OpenAI have had security missteps in the past, leading some to question whether the service will be considered a safe option by US government agencies.