Delivering Cutting-Edge AI Solutions to US Government

Candice Ling

October 31st, 20232 19

Artificial intelligence (AI) has the potential to transform nearly every sector of the economy and society. For U.S. government agencies, the latest advances in AI offer a unique opportunity to optimize their operations, deliver better user experiences and enhance national security. Microsoft is innovating in AI to deliver technology for the public sector. As we embark on this new way to work, we are committed to delivering transformational AI innovations responsibly to drive government missions, aligned to the goals of the White House’s recent executive order for safe, secure and trustworthy AI.

We began enabling government agencies to adopt new AI technologies for mission critical solutions with Azure OpenAI Service in the Azure commercial cloud. Azure OpenAI Service is included as a service within the FedRAMP High authorization for our commercial cloud, enabling many agencies to securely access new large language models (LLMs) like GPT 3.5 and GPT 4 for their unclassified (non-CUI) workloads.

However, we recognize that many agencies handle sensitive data requiring higher levels of security and compliance. To enable these agencies to fully realize the potential of AI, over the coming months Microsoft will begin rolling out new AI capabilities and infrastructure solutions across both our Azure commercial and Azure Government environments to support critical government needs and further drive innovation for mission critical workloads. These offerings include:

Azure OpenAI Service in Azure Government

In the first quarter of calendar year 2024, Microsoft plans to launch an initial preview of Azure OpenAI Service in Azure Government for government agencies. This will enable the latest generative AI capabilities, including GPT-3.5 Turbo and GPT-4 models, for customers requiring higher levels of compliance and isolation.

Azure OpenAI Service for Azure Government is purpose-built, AI-optimized infrastructure providing access to OpenAI’s advanced generative models that can help government agencies improve efficiency, enhance productivity, and unlock new insights from their data. Azure OpenAI Service enables a wide variety of use cases to help government users support their unique missions, including:

Content generation: Creating a realistic output, including textual and visual content, based on input data or natural language prompts
Content summarization: Extracting key themes and insights from a longer piece of text, including answering natural language queries
Semantic search: Going beyond traditional keyword matching by understanding the meaning behind a query and retrieving relevant search results that are semantically related to the user’s intent
Code generation: Generate computer code based on a prompt, translating code from one programming language to another and reviewing and improving existing code

Azure OpenAI Service for classified workloads

Expected during the summer of 2024, Microsoft plans to preview Azure OpenAI Service in our air-gapped classified clouds to select national security customers. This solution brings a distinct generative AI platform in our isolated classified cloud environment, enabling national security leaders and operators to use critical AI capabilities to analyze and action on highly sensitive data anytime and anywhere. National security customers working in classified environments will also have access to open source solutions, via Azure Machine Learning, as early as the first quarter of calendar year 2024 to complement our expanded Azure OpenAI Service in our classified clouds.

Teams Premium with intelligent recap in Microsoft 365 Government

Teams Premium with intelligent recap is expected to begin roll out to government users in GCC during the Spring 2024. Intelligent recap in Teams Premium uses AI to help users focus on the meeting content that matters most. AI-generated meeting notes and tasks help users save time catching up and coordinating next steps. Automatically generated chapters and topics make it easy for users to navigate recordings and identify the most relevant sections. Availability details for Teams Premium with intelligent recap in GCC High and DOD are expected to be shared in early 2024.

Microsoft 365 Copilot update for Microsoft 365 Government

Microsoft 365 Copilot GCC is expected to begin roll out to government users in the GCC environment during the Summer of 2024. At launch, agencies will have access to this transformational AI assistant in GCC, bringing generative AI to our comprehensive productivity suite for a host of government users. The Microsoft 365 Copilot GCC offering will meet security and compliance requirements for U.S. federal, state, local and tribal government agencies. Availability details for Microsoft 365 Copilot GCC High and DOD are expected to be shared in early 2024.

Microsoft Copilot combines the power of large language models (LLMs) with your data in the Microsoft Graph, the Microsoft 365 apps, and the web to turn your words into the most powerful productivity tool on the planet. Microsoft 365 Copilot offers Microsoft’s comprehensive approach to security, compliance, privacy, and responsible AI.

Microsoft 365 Copilot supports productivity across workloads by helping you quickly catch up on meetings you missed or summarize a meeting in real-time. Instantly uncover key insights across all your data. And power through your stack of emails with instant replies with a personalized length and tone. Agencies can take steps today to ensure Microsoft 365 Copilot readiness.

Open Source LLMs in Azure Government

Azure Machine Learning comes with the ability to train, deploy, and manage machine learning models, including open-sourced foundation models. While Microsoft is committed to providing industry-leading large language models from OpenAI in Azure Government, we also recognize that some mission requirements benefit from smaller generative AI models, such as Llama 2. As of today, Microsoft has enabled access to Llama 2 via the Azure AI model catalog in Azure Government. Azure Machine Learning will enable customers to envision, develop, and iteratively improve business-critical machine learning models at scale.

Building a Partner Ecosystem to Foster and Scale AI Innovation

In addition to our first party offerings, Microsoft is committed to growing a robust partner ecosystem to help drive further innovation. Partners are key to how we build and scale innovation for our customers. In this new era of AI, we continue to rely on expanding partnerships with creative companies and integrators to help us deliver AI at scale, securely.

One example is our early and successful collaboration with Ask Sage, with whom we are delivering a secure and extensible generative AI platform for government and industry. Ask Sage is home to several thousands of government and partner teams that are using the power of Azure OpenAI and their plugin architecture to bring a range of transformative innovations that keep humans in the loop while leveraging AI to work seamlessly across systems and data silos.

Another example is a new collaboration with Mattermost and MobiusLogic. Together, we are working to deliver improved collaboration and workflow solutions for national security customers built on Azure’s secure cloud and leveraging Azure AI capabilities. The solution stack is built on the Mattermost OpenOps AI experience platform connected across Microsoft’s core platform services including Azure AI Services, Microsoft Entra ID, and Microsoft Teams. The hub offers a highly interoperable collaboration experience with advanced levels of customization, integration, AI-acceleration and deployment across private cloud, public cloud and air-gapped networks.

Our commitment to developing and deploying AI responsibly

At Microsoft, we’re committed to building and deploying AI technology in a responsible and ethical manner. Generative models such as the ones available in Azure OpenAI Service and Microsoft Copilot have the potential for significant benefits, but without careful design and thoughtful mitigations, such models could generate incorrect or even harmful content. Microsoft has made significant investments to help guard against abuse and unintended harm, which includes, but is not limited to, requiring applicants to show well-defined use cases, ensuring our engineers follow the Responsible AI approach, building content filters to support customer needs, and providing responsible AI implementation guidance to onboarded customers.

As we work in support of the public sector’s missions, it’s crucial to understand the security, privacy, and compliance measures of the Azure OpenAI Service and Microsoft Copilot. Microsoft ensures that your data is processed, used, and stored in accordance with the Microsoft Products and Services Data Protection Addendum.

In addition, Microsoft has established the Responsible AI standard which provides a framework for developing and using AI systems responsibly. Impact Assessments are conducted early in the system’s development to assess potential impacts on people, organizations, and society. Oversight mechanisms are implemented to identify AI systems that may have a significant adverse impact. Customers are encouraged to leverage these standards by conducting Impact Assessments early in the development of AI systems, implementing oversight mechanisms for systems with potential adverse impacts, and updating Impact Assessments regularly.

Stay up to date

Interested in receiving more news and announcements on AI in government? Sign up here to get timely updates delivered to your inbox.