Microsoft has offered large language models (LLMs) developed by OpenAI to US federal agencies using Azure cloud services. The company announced this on the official blog.
According to Reuters, Microsoft has added support for both GPT-3 and the newest and most complex LLM variants behind the GPT-4 multimodal model in the Azure Government cloud.
There has been a boom in the use of LLM since OpenAI, backed by Microsoft and largely controlled by the latter, introduced a variant of ChatGPT with never-before-seen characteristics of “sociality” – a chatbot is able to maintain a completely “meaningful” dialogue, and now businesses of all sizes – from startups to large corporations, they are trying to present comparable solutions and similar functions in their services.
Microsoft is the first to provide GPT technology in the Azure Government cloud, which provides services to many government departments in the United States – this is the first known attempt by large companies to provide a service of this kind to government agencies.
However, Microsoft is already offering this functionality to commercial Azure users by providing access to Azure OpenAI Services. As of May 2023, there were already 4.5 thousand customers.
Microsoft said that government clients can give the AI a variety of tasks, including generating content, converting text commands into program code, or, for example, compiling summaries of voluminous materials.
It is worth noting that LLM-based services can still suffer from “hallucinations” and give frankly wrong answers, and experts and industry leaders have repeatedly warned about the dangers of AI and the need to regulate such developments and solutions, and even called their products a threat comparable to a pandemic and nuclear war.
If you notice an error, select it with the mouse and press CTRL + ENTER.