• Constraints imposed by an austere and remote operating environment. 
• Increased cognitive load on individuals conducting operations due to exponential growth in the volume, veracity, and velocity of data. 
• Survivability and the need for distributed nodal command and control.

The dilemma posed here is whether technological advancements inadvertently compromise decision-making abilities due to the heightened cognitive burden on users. 

Decisive action powered by AI 

Speed, precision, and data are critical on the modern digital battlefield. Human-machine teaming allows modern soldiers to work with AI as their digital agents, using natural language or voice commands through military radios. This hands-free interaction improves situational awareness and enhances decision-making by combining AI’s analytical power with human intuition and judgment.  

Using AI and machine learning on missions will become critical to effective command and control environments. Language models have evolved to create and use enterprise-level knowledge bases, integrating external data for more complex interactions. This advancement has significant effects for mission capabilities, with early applications in: 

• Voice transcription and translation—We have already seen that when paired with Push-To-Talk (PTT) voice radios, digital audio voice streams can be captured for real-time transcription, translations, and augmentation with other sources of data. 
• Robotic command and control orchestration—With an intent to release operators from the need to operate these systems manually, we can not only free human resources to concentrate on the specifics of their mission but also reduce the force protection overhead that is required to keep operators safe. 
• Intelligence, Surveillance, and Reconnaissance (ISR) analysis—Working with multiple agents and multimodal sensors for defense use cases, we can help increase the accuracy and range of surveillance and provide a multilayered approach to detection and action. 
• Querying Battle Management Systems—We not only provide the capability to access information in a humanistic way, at the point of need, we also reduce the intense staff effort associated with briefing and analysis of data—the AI agents can take on the manual load, freeing up the human cognitive load to enable better and faster decision making. 

Agentic AI explained 

So, what do we understand about the advancement and application of Agentic AI? When discussing Agentic AI, it’s crucial to highlight the characteristics that distinguish an agent from tools like ChatGPT or traditional digital assistants we’ve seen in office settings. There are five key nonlinear elements that define agentic capabilities: 

• Planning—Instead of diving right into a task, an AI agent pauses and plans the series of steps required. This structured approach prevents errors, as we often see in traditional language model implementations with robots. 
• Reflection—Current models like ChatGPT provide answers but don’t validate them, as they lack a built-in ‘reflection’ capability. The ability to ‘reflect’ and ensure completeness is crucial to confirm that tasks are executed properly and are relevant to each subsequent step in the Agentic AI lifecycle. 
• Use of tools—When the AI agent encounters a step it can’t perform, it checks its manual for a corresponding tool, gathers needed information, executes the task, and processes the response. This is crucial for proprietary industry capabilities, allowing handoffs to external sources. 
• Collaboration—Where the human or agents work collaboratively on collective tasks. This is important for two reasons: creating clear boundaries and ensuring agents are task-specific.  
• Memory—This cycle is further powered by memory, where the agent retains and can recall prior inputs, actions, and outcomes. With this memory, the agent learns from past decisions, allowing it to improve future actions and refine its planning and reflection. 

Traditional non-agentic AI workflows vs agentic AI workflows 

Collectively, these five characteristics form a framework known as the REACT framework (Reasoning and Action). Reasoning involves planning and reflection, while action is about the execution.  

The key difference between traditional non-agentic AI workflows, often seen in zero-shot prompts, and the more advanced, agentic workflows we’ve been discussing can be seen in the diagram below. 

In practice, AI agents can be seamlessly integrated into an organization’s workflow, especially for field operators. This will result in more efficient missions, quicker responses, and a trusted pairing of humans and machines. Additionally, it will allow warfighters to focus on tactical operations while AI handles data processing and situational analysis in the background.  

This is where digital agents can come into play. Digital agents that allow operators, particularly those in forward positions, to delegate specific tasks using natural language. Incorporating these agents into your workflow can help revolutionize how your organization handles complex operations. By offering an intuitive interface, robust performance under duress, and the ability to manage tedious tasks, these agents ensure that operators at the tactical edge can focus on what really matters—making critical decisions in dynamic environments. 

Microsoft AI principles 

Microsoft is committed to advancing AI through principles that put people first. 

We put our responsible AI principles into practice through the AI, Ethics, and Effects in Engineering and Research (Aether) Committee, as well as our Office of Responsible AI (ORA). The Aether Committee advises our leadership on the challenges and opportunities presented by AI innovations. ORA sets our rules and governance processes, working closely with teams across the company to support the effort. 

Microsoft AI serves to enhance human capabilities, not replace them. It’s designed to embody principles such as fairness, inclusivity, reliability and safety, transparency, privacy and security, and accountability. By using AI to optimize administrative functions and services, stakeholders can focus on what matters most: human-centered design, decision-making, and empathy.  

Implement emerging technologies strategically 

Defense decision makers should consider not just what AI can do, but what it should do to innovate in a reliable and trusted way. It’s critical to understand the components of a holistic approach to AI that will help agencies turn meaningful innovation into actionable results that will benefit society.  To learn more contact your Microsoft Defense and Intelligence representative today, or engage with the following Microsoft resources:  

• Microsoft Agentic Agents.
• AutoGen Update: Complex Tasks and Agents.
• Copilot agent builder.  

The post Empowering defense operations with Microsoft AI appeared first on Microsoft Industry Blogs.

Microsoft Cloud for Sovereignty is at the forefront of this movement, offering comprehensive solutions tailored to diverse regulatory landscapes, enabling organizations to streamline their compliance processes and meet necessary standards efficiently and effectively while in the hyperscale cloud. The most recent updates are designed to help customers navigate the complexities of regulatory frameworks, streamline compliance processes, and enhance their overall data control. By leveraging these new capabilities, organizations can confidently advance their digital transformation while maintaining the highest standards of compliance and data protection.

Microsoft Cloud for Sovereignty

Learn more

The latest from Microsoft Cloud for Sovereignty

Microsoft Cloud for Sovereignty is introducing several significant updates:

1. Policy portfolio enhancements: The policy portfolio now includes new initiatives like Network and Information Security Directive (NIS2) and others to help regulated industries in their compliance requirements.
2. Sovereign Landing Zone in Terraform: The Sovereign Landing Zone (SLZ) can now be deployed using Terraform, providing advanced sovereignty controls and capabilities supporting regulatory compliance for organizations with stringent regulatory requirements. 
3. PubSec Information Assistant now has secure mode, which helps organizations meet their sovereignty objectives and address data protection and privacy issues. 
4. Guidelines for Azure Databricks in regulated industry: Azure Databricks guidelines and reference architecture help users build a well-architected environment that adheres to best practices for regulated industries. 
5. Regulated Environment Management portal (REM) now in private preview, offers a suite of Azure services to manage the entire lifecycle of cloud environments, helping meet compliance and transparency needs for industries including public sector and beyond.  

Policy portfolio enhancements

The Microsoft Cloud for Sovereignty policy portfolio, which expands on the Azure built-in policy initiatives, helps customize deployments to align with specific custom policy frameworks. A policy initiative is a collection of technical controls mapped to regulations and frameworks published by governments or organizations. By leveraging these policy initiatives, organizations can streamline their compliance processes, reduce time needed to audit environments, and help meet both established regulatory compliance and government requirements efficiently and effectively.  

Currently, the portfolio includes six publicly available initiatives for varying data classification levels, with more in development. Notable examples of these initiatives included in this latest update are the Network and Information Security Directive 2 (NIS2), Spain Esquema Nacional de Seguridad (ENS), and New Zealand Information Security Manual (ISM). 

We continue working to enhance the Azure policy initiative for NATO’s D32 directive on the protection of information in the public cloud to support diverse security requirements across levels of classification for the protected business network in a public cloud at NATO. A policy initiative preview has been developed for the NATO Unclassified requirements, and efforts are underway to create a policy initiative for the NATO Restricted requirements.

Additional information and details about the policy initiatives is available in our Microsoft Cloud for Sovereignty policy portfolio GitHub repository. 

Sovereign Landing Zone in Terraform

In addition to existing Bicep deployment support, Terraform support is in public preview. The SLZ enables organizations with advanced sovereignty needs to address their regulatory compliance requirements using Terraform, the industry-standard multi-cloud tool for Infrastructure-as-Code (IaC) and Policy-as-Code (PaC) capabilities. Customers can now use Terraform to configure and deploy their landing zones and enforce compliance effectively—underscoring our commitment to providing robust and compliance-oriented cloud solutions tailored to help customers meet the stringent regulatory demands of today’s digital landscape. 

AI in the public sector

Public sector organizations often face the challenge of managing vast amounts of data and require an intelligent solution to quickly find relevant, accurate answers with source references, all in a secure manner. The Information Assistant agent now available on GitHub, can enhance employee productivity by significantly reducing the time needed to locate crucial information within an organization’s collective knowledge base. 

This AI solution serves as a starting point for organizations to develop their own custom generative AI capabilities, leveraging the power of Azure OpenAI Service. It demonstrates a common scenario where large language models (LLMs) are used to “chat with your own data” through the retrieval augmented generation (RAG) pattern. This pattern allows LLMs to generate responses based on your domain data without the need for model fine-tuning. Specifically, this solution is now compatible with Microsoft Cloud for Sovereignty, deployable within the Sovereign Landing Zone (SLZ) online management group with secure mode, and compliant with the Azure policy built-in sovereignty baseline global policy initiative to enforce data residency requirements.  

This update introduces secure mode, a significant step forward in assisting organizations to meet their sovereignty objectives and address data protection and privacy issues. This feature aims to safeguard sensitive data by employing security measures such as encryption and network controls to block unauthorized access. By simplifying many intricate and time-consuming tasks associated with securing systems, secure mode helps organizations bolster their security stance. This not only saves time that would otherwise be spent on manual setups and verifications but also enhances the consistency and dependability of the security measures implemented. 

Microsoft is committed to providing guidance and workload accelerators that empower customers with sovereign requirements to build their own generative AI and agent solutions, leveraging the hyperscale cloud.

Guidance for Azure Databricks in regulated industry

The newly published reference architecture and documentation is available to provide valuable guidance on using Azure Databricks within a Sovereign Landing Zone or Azure Landing Zone deployment. Azure Databricks has several features to help secure environments address stringent regulatory requirements, providing robust security measures and compliance capabilities. It supports data engineering, data science, and data analytics workloads with interactive notebooks for coding and visualization, various compute options for scalable and efficient processing, and seamless data integration with multiple Azure services.

Regulated Environment Management portal

The Regulated Environment Management (REM) portal, currently in private preview, plays a crucial role in Microsoft Cloud for Sovereignty by catering to the public sector and heavily regulated industries. Starting November 2024, REM will transition from a tenant-based to a subscription-based service. This transition brings several new benefits, including configuration data isolation and regional boundaries, managing hundreds of landing zones as a group, including managing permissions on the group level. 

We are also introducing the Landing Zone Account (LZA) to serve as a group for creating, configuring, deploying and managing landing zones—all from a single place. Each landing zone account is associated with a region and a subscription. LZA offers data isolation and residency, simplifying the management of landing zones, while still maintaining the flexibility to deploy the landing zone in any region. For developers, REM now includes a representational state transfer (REST) application programming interface (API), enabling flexibility and ease of integration.

Transparency logs are available as a preview Azure service, providing customers with visibility into occasions when Microsoft engineers accessed customer resources using the just-in-time access service. Customers can now configure transparency logs directly through the REM portal via an Azure resource, simplifying the onboarding experience. In addition to these capabilities, REM allows customers to manage their policy portfolios through the Azure portal, including deploying policy initiatives to Azure. 

Moving forward

Implementing Microsoft Cloud for Sovereignty capabilities is pivotal in enabling organizations in government, the public sector, and heavily regulated industries to meet their compliance requirements while leveraging the latest digital innovations. By staying committed to transparency, security, and compliance, Microsoft Cloud for Sovereignty will empower organizations to navigate the complexities of the digital age with confidence and agility.

Learn more about Microsoft Cloud for Sovereignty

• Microsoft Cloud for Sovereignty 
• Microsoft Cloud for Sovereignty product documentation 
• Sovereign Landing Zone (GitHub) 
• Policy portfolio (GitHub) 
• Quickstarts (GitHub) 
• Public Sector Center of Expertise and Digital Skills

The post Accelerate cloud adoption with Microsoft Cloud for Sovereignty appeared first on Microsoft Industry Blogs.

As Microsoft’s global government lead, I am inspired that so many agencies and organizations are not hedging their bets with AI, as was sometimes the case with prior waves (cloud computing and the rise of mobility come to mind). Instead, many governments are taking thoughtful early steps to explore AI and invest in early innovation.

Microsoft for Government

Achieve your mission with Microsoft.

Learn more

How AI is transforming government IT

At Microsoft for Government, our job is to help governments make the most of their AI investments with best-in-class solutions that are tailored to their unique needs and brought to life through our global network of technology partners. Since the advent of generative AI in late 2022, we have worked closely with a wide array of agencies and organizations to develop targeted use cases designed to improve everything from citizen services to cybersecurity.

This level of innovation is even more impressive considering the unique pressures facing governments. Leaders must deliver results within constrained budgets and limited resources, respond to shifts in societies and workforces, and protect some of the most heavily cyber-attacked data and systems in the world.

Here’s a brief look at four essential areas where AI is helping to empower customers and reshape the impact of government around the world.

1. Delivering personalized experiences

Unproductive interactions with automated call centers and outdated websites are the bane of many constituents’ relationships with their government agencies. Well-intended at the core, too many of these services fail to connect with constituents, recognize their needs, and deliver useful information in a timely way.

Generative AI is revolutionizing these resources with self-service portals and contact centers that provide modern, user-friendly digital experiences. The combination of natural language processing (which enables machines to understand and generate human language), and semantic search (which goes beyond simple keyword matching to comprehend the meaning of a query) makes it possible to readily create chatbots and agents that interact smoothly with people and reason over vast amounts of data to instantly provide solutions.

An amazing example is a chatbot in India called Jugalbandi, built on Microsoft Azure OpenAI Service. Jugalbandi has bridged the nation’s vast linguistic divide by understanding spoken and typed questions, then responding in the user’s local language, all on widely used mobile devices. By covering 10 of India’s 22 official languages and 171 government programs (so far), it makes important information more accessible and reduces cost and friction for constituents and employees alike.

2. Empowering the government workforce

Ensuring that employees can focus on meaningful work is key to government success. In public service, many people are motivated by the opportunity to help others and solve important societal challenges. When subpar technology solutions create inefficiencies and cause unintended problems, productivity and morale suffer.

Government workers, like employees everywhere, welcome services and solutions that reduce friction, keep them connected, and make them more productive. Solutions that use generative AI’s unique predictive analytics and forecasting capabilities can help employees improve their decision-making and overall efficiency.

In the United Arab Emirates, the Dubai Electricity and Water Authority (DEWA) DWEA) built a new Business Requirement Document Generator using Microsoft Azure AI Services that enables employees to enter minimal information and easily generate important documents. Trained on 500 billion parameters, it proactively suggests required information and fills in gaps and other important information. What used to take one week can now be done in one day, including reviews. Complementing the solution, the Authority also boosted productivity by adopting Microsoft 365 Copilot, which is proving invaluable in helping employees to work faster and effectively, notably on research tasks.

3. Modernizing government operations

In many organizations and agencies, the journey to cloud computing has been hindered by important considerations unique to government, including cost, security, and a reliance on legacy systems which are often regarded as so critical that they can’t be modernized, despite being slow, complex, and vulnerable. As cloud and AI solutions become more mature and available, however, the barriers to modernization are dropping, while the price of inaction only rises.

Generative AI’s code generation capabilities (which empower non-developers to readily create AI applications) make it possible to manage critical workloads in the cloud, while gaining new benefits in flexibility, scalability, and resilience across agencies. This can unlock critical data stuck in silos and drive better decision-making for officials who, for example, can make more informed decisions on proposed legislation. Likewise, for citizens, it can turn the process of obtaining a building permit into a productive interaction with a helpful virtual agent, rather than a gauntlet of arcane rules and regulatory snags.

When Gamle Oslo realized that a district manager needed to log into 25 different systems to find relevant data for the city’s kindergarten services, they decided to build a unified data platform using Microsoft Fabric. The solution collects and analyzes all of the district’s data on housing, employment, health care, and public services, which had formerly been disconnected. Unified data with Fabric has improved many processes and opened the door to new benefits and insights, including setting the stage for new AI innovation.

4. Securing data and protecting resources

More than any other industry sector, governments are prime targets for cyberattacks. According to Microsoft’s Digital Defense Report for 2023, 53% of attacks worldwide focused on government organizations and critical infrastructure². Over the last 18 months, public sector organizations have seen a 150% increase in cyberattacks³ due to the combination of escalating geo-political conflicts and increasing financial motivations.

This dangerous threat landscape urgently requires governments to improve their ability to safeguard critical systems, enhance data protection, and maintain compliance with a host of regulations. Making the challenge even more difficult for governments is the growing demand for skilled cybersecurity staff. Worldwide, there is a shortage of 4 million cybersecurity professionals⁴, and the problem can be worse for governments, who often struggle to attract and retain talent.

Cloud and AI technology can help level the cybersecurity playing field in a number of important ways. First, the Microsoft Cloud platform provides built-in security and resilience, reflecting our commitment to making security and compliance our top priority. Also, our comprehensive security offerings help an organization craft its own strategy using end-to-end products and tools. And finally, Microsoft Copilot for Security uses generative AI to empower security analysts to rapidly assess an organization’s security posture and create actionable insights and solutions at much greater speed than current approaches.

When the Dominican Republic’s National Cybersecurity Center (CNCS) recognized increases in both frequency and sophistication of cyberattacks, they responded by modernizing the government’s information system and in national critical infrastructure. Their comprehensive approach included adopting a variety of Microsoft solutions to monitor, analyze, manage, and respond to cyber threat cases across the country. The benefits of modern cybersecurity were soon evident, as the event correlation system running on Microsoft Sentinel established relationships between indicators of compromise four times faster than earlier approaches.

See AI in government for yourself

Beyond these transformational benefits, the explosion of AI innovation in government is also advancing the move to modern cloud and data platforms, which is not only essential for enabling AI but also delivers broader benefits in efficiency, productivity, and security.

The insights and examples shared here only just scratch the surface of what our customers are creating and learning. We’re discovering new and amazing things on a daily basis, and we invite you to become engaged.

To see for yourself and learn more:

• Attend the Microsoft AI Tour—a free, one-day event with experts, industry leaders, and peers to explore how AI can drive growth and create lasting value. Events are happening worldwide through March 2025.
• Visit us at Smart City Expo World Congress in Barcelona, Spain, November 5–7, 2024. I will present on the topic of “Next-Gen Tech Leading the New Urban Era​,” and other Microsoft for Government sessions will be presented by Kathleen Mitford, Kirk Arthur, Jeremy Goldberg, Doug Priest, and Hannah Prior.
• To discover how the city of Madrid is innovating to become a smarter, more sustainable city, download the new SmartCitiesWorld City Profile. And for a discussion of the new profile, including a new AI-driven virtual assistant, register for a SmartCitiesWorld webinar on October 22, 2024.

Sources:

¹Gartner, Compare AI Software Spending in the Government Industry, 2023-2027, By Daniel Snyder, James Ingham, Inna Agamirzian, 27 March 2024. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.

²Microsoft Digital Defense Report 2023

³Statista, December 2023

⁴ISC2 Cybersecurity Workforce Study, November, 2023

The post 4 foundational ways that AI is transforming government appeared first on Microsoft Industry Blogs.

Governments do remarkable work to help people live with dignity and endure difficult circumstances by providing essential benefits and resources across a broad spectrum of programs. Yet the challenge of providing the right benefits at the right time to all qualified recipients while also minimizing the incidence of improper payments is a balancing act that many governments struggle to achieve. 

Helping agencies and organizations to address these types of challenges is central to our work in Microsoft for Government. We focus on enabling thriving communities and inclusive programs through technology. In the case of benefits protection, we work to help each government discover the right mix of technological and organizational innovation to mitigate risk within the parameters of their unique circumstances and requirements. 

Microsoft for Government

Help solve society’s biggest challenges.

Learn more

The rising tide of improper payments

The loss of benefits funds to unauthorized payments has long been a problem for governments around the world. In recent years, the scale and complexity of the challenge has escalated. Fraud is a particularly expensive and multidimensional problem. To offer just one example, the US Government Accountability Office (GAO) estimated that the amount of fraud in US unemployment insurance programs during the COVID-19 pandemic was likely between $100 billion and $135 billion USD.¹

The reasons for this are varied, and they parallel larger trends in society. A key driver is the ready availability of AI tools, which fraudsters have enthusiastically embraced to automate, scale, and generally uplevel the impact of their efforts. But it’s not always a matter of criminal intent. So-called “unintentional errors”—mistakes made without maliciousness due to things like administrative errors or poorly designed solutions—also incur huge costs. According to the GAO, overpayments or payments that should not have been made (for example, to deceased people or those no longer eligible for government programs) totaled an estimated $247 billion during fiscal year 2022.²

Why governments struggle to ensure proper payments 

Unlike organizations in other global industry sectors, governments face unique challenges in administering complicated social benefits programs. The nature of distributing funds accurately and efficiently to constituents in all corners of society is inherently fraught. Governments must contend with a set of vexing factors, including: 

• Growing demand for benefits funding—A mix of aging populations, increases in immigrant and refugee populations, and people seeking broader ranges of benefits is driving higher budgets and greater complexity. 
• Mandates to “do more with less”—Austerity programs and budget cuts are colliding with rising public expectations for modern services and impacting the ability of security teams to procure top cybersecurity talent.  
• Disconnected data silos—Across departments and agencies, teams often work with older tools and isolated systems that cannot share data easily or securely, making integrated new “tell us once” solutions difficult to implement. 
• Ineffective legacy systems—A high “technical debt” of outdated computing environments and applications impedes innovation and threatens to hinder the most vulnerable people from accessing payments they’re entitled to.  

Beyond these costs and considerations, the larger risk here is the erosion of public trust. Especially in social benefits systems, which exist to ensure the welfare of people across all walks of life, government is obligated to demonstrate reliability and integrity. Widespread fraud and abuse do more than increase deficits and siphon off critical resources. They also undermine confidence in all aspects of government.  

A vision for modern benefits delivery and protection 

Imagine if governments were able to innovate on par with private industry. Imagine if people could easily access all the information they need through any channel (online, over phone, text, and more) and have their own virtual assistant to help them apply for benefits using natural, everyday language. Imagine if these systems would inherently enforce rules and keep up with constituents’ changing lives. Imagine if customer service representatives had access to a complete, 360-degree view of the constituent—if it knew their history and all their previous contacts with every social service agency. And imagine that such things could be accomplished without breaking the budget.  

These are obviously lofty, aspirational visions. However, thanks to the relatively recent technological advancements, some innovative government social service organizations are doing just that. The advent of generative AI and the innovation it has inspired has already delivered promising results on many fronts of government operations, including benefits. To cite just one example, a generative AI-powered chatbot developed in India called Jugalbandi is helping people get assistance for any of 171 government programs, simply by conversing through mobile devices in 10 of India’s 22 official languages. “This is revolutionary for people who could not interact with tech because of language barriers,” said Abhigyan Raman, a project officer. 

Ensuring the integrity of benefits payments means reevaluating a government’s existing slate of technology investments and charting a multistep course of transformation. Embracing AI is a common theme, but the destination in terms of functional outcomes and benefits will be unique to the organization.  

Three areas of generative AI impact in benefit payments protection

In our work with customers who are either early in the journey or simply just embarking, we identify three categories of solutions that hold the greatest promise for cloud and AI innovation:

1. Boost workforce productivity. Governments can potentially save time and money using generative AI for scenarios such as live transcriptions and translations during benefit eligibility interviews. AI can also draw insights more easily from large volumes of data. One department, for example, identified GBP14 million in potentially fraudulent loans by analyzing a set of 250 networks of people, organizations, and places, processing 100 million data items.

2. Embrace “Prevention by Design.” AI can integrate controls into systems to detect fraudulent or erroneous activities and enable real-time profiling and alerts. For example, a leading bank is using Voice ID analysis for incoming phone calls from 31 countries and 15 languages, checking more than 100 behavioral and physical vocal traits in a matter of seconds. Since launch, it has flagged more than 43,000 fraudulent calls and reduced fraud by 50%, preventing an estimated GBP981 million in losses.

3. Enhance citizen engagement. When chatbots become knowledgeable assistants, when a person’s voice becomes their password, when the disabled have equal access to resources, then everyone will benefit. For example, a government department is developing an Intelligent Website AI Assistant to help taxpayers more easily process their tax returns. It uses natural, everyday language to address queries and disseminate timely and consistent information. This will help improve compliance and avoid revenue leakages from fraudulent activities.

Beyond technology: How Microsoft partners for the long term 

Embracing the paradigm shift of generative AI obviously begins with technology. The table stakes of modernization for government include migrating to a modern cloud platform and the adoption of a comprehensive AI development solution from a vendor who demonstrates a deep commitment to security and responsible AI practices. 

Microsoft invests heavily in all these areas. However, success involves much more than just technology. Governments also depend heavily on the contributions of trusted solution providers, and we believe our global partner ecosystem sets us apart, with expertise in all corners of the world. Then, the final unique benefit we offer is the deep experience of our industry advisors and the many highly experienced government veterans on the Microsoft Government team. 

Our job is to help build the bridge between the technical and the strategic, on realistic terms. When we sit down with customers, we help clarify challenges and goals, educate on important challenges (for example, how governments can tackle cybersecurity and AI skilling), and share our experiences with other governments facing similar challenges (sometimes even connecting them to help foster learning). Then, we embark on identifying and exploring use cases, evaluating impact, and taking the knowledge gained for further innovation.  

We are excited to work with governments to mitigate fraud and abuse in payments. To learn more about how Microsoft is helping to create opportunities that support vulnerable communities, see our Microsoft public health and social services website, and learn more about Microsoft for Government. 

¹Estimated Amount of Fraud during Pandemic Likely Between $100 Billion and $135 Billion, GAO. September 2023.

²Federal Payment Errors, Known As Improper Payments, Are A Continuing Concern, GAO. March 29, 2023.

The post How Microsoft and AI work to reduce government benefit fraud and error appeared first on Microsoft Industry Blogs.

They decided to migrate their 10-year-old architecture and technology stack to a cloud-based solution to support their long-term goals and operate more efficiently. The question was how to make the move while also keeping the power on.  

At Microsoft for Government, we work with organizations and agencies around the world to help solve these kinds of challenges. We help each organization navigate their unique requirements and chart a path that works best for them, balancing the promise of AI and cloud native applications with the need to be more cost efficient, secure, and compliant.  

For Energinet, the solution was to build a new digital operating system based on Microsoft Azure, which is improving efficiency in automating energy balancing processes, lowering costs, and resolving issues in 15 minutes that previously took an hour. Modernizing helped them meet their near-term requirements, and it positioned them to remain agile and open to change as opportunities evolve.  

Azure Migrate and Modernize and Azure Innovate

Get guidance, browse resources, and find expert help for your move to Azure

Discover the benefits

Migrate to innovate—the benefits of modernization

In many government organizations, the migration to cloud computing has long been mitigated by important concerns, including cost, security, and a need to maximize legacy systems. As cloud technology has matured, however, the barriers to adoption have declined as the price of inaction has risen.  

With the promise of AI to deliver new efficiencies and opportunities for service improvements, cloud migration is now clearly the best path forward, provided it is done in ways that meet an organization’s unique requirements.  

For government organization, the advantages of cloud computing are compelling, and easier to realize than before. Governments that choose to migrate to Azure are realizing important benefits in many areas of their greatest concerns—among them: 

• Performance and resilience: Azure enables businesses to scale their operations globally with ease, with purpose-built infrastructure, dynamic compute capacity, scalable storage, and real-time disaster-recovery options. Governments such as the State of Alaska are migrating to Azure to achieve their vision of becoming digital public service innovators. To expand access to secure services, Alaska migrated 700 applications and one-third of its infrastructure in just three months. They not only achieved better resilience, cost efficiency, and security, but with the state’s vast geography and often isolated communities, the migration also sparked a cultural shift by bringing agencies together and unlocking unexpected value. 

• Security: To counter the expanding cybersecurity threat facing governments, Azure is supported by more than 8,500 security experts, more than 100 compliance certifications, and a cloud-native application protection platform that spans the application and infrastructure stack. Microsoft plans to invest USD$20 billion in security in the next five years to continue our commitment to a safe future.¹ The importance of ensuring world-class security was a key factor in the decision by Qatar’s Ministry of Communications and Information Technology to digitize government operations with Azure. The ministry established information assurance measures and security programs that not only enhanced the government’s data security and operational efficiency, but also achieved $7.3 million in cost savings. 

• Hybrid and multi-cloud management: Azure supports hybrid, multi-cloud, and edge environments with Microsoft Azure Arc, a solution that allows governments to build applications and services with a consistent development, operations, and security model across deployments. This proved essential for the World Bank, which provides lending services in 189 developing countries around the world to help lift people out of poverty. They wanted to build applications and services that could extract insights from their SQL Server estate and multiple cloud infrastructure providers. Using Azure Arc, they streamlined their cloud migration journey, built new solutions, and gained unexpected efficiencies. 

• Cost savings: By migrating to Azure, customers can optimize costs and resources by consolidating solutions and choosing from a variety of consumption models and flexible migration approaches. For example, the Statistical Office of Republic of Serbia saved time and money by conducting the nation’s first-ever paperless census using a hybrid cloud solution, which reduced the time required to publish official results from 18 months to just six. The solution delivered faster data encoding and more accurate results, and the Statistical Office of Republic of Serbia was able to streamline maintenance while ensuring optimal security, real-time monitoring, and improved data quality.  

A 3-step approach to becoming AI-ready

Modernization, which provides for greater scale, efficiency, and flexibility, also positions an organization to explore the benefits of AI. Governments recognize the potential of AI to generate new cost efficiencies and to power new offerings in service delivery. In the near term, this is motivating many to accelerate their digital transformation journeys. 

Implementing a cloud migration strategy is an absolute prerequisite to adopting and innovating with AI in a government organization. The cloud provides the hyperscale performance required for generative AI functionality, and a modern data strategy not only consolidates disparate data systems but also ensures access control and data security.  

Cloud migration is a long-term process, and the journey is unique to every organization. Whatever the course, governments should remain mindful of the following three steps, which are key to becoming AI-ready in ways that are efficient, effective, and responsible. 

Step 1: Co-locate data and workloads in the cloud

Strategically placing applications, databases, and AI resources into the Microsoft Cloud ecosystem delivers exponential improvements in performance and prepares data and services to take advantage of new AI innovations.  

Step 2: Infuse Microsoft Azure OpenAI Service and copilot integrations

Once your data and applications are co-located, you are ready to take advantage of AI services such as Microsoft Copilot for Microsoft 365, which integrates generative AI into everyday productivity applications, and Azure OpenAI, which enables the development of customized copilots, plugins, integrated AI services, and much more.  

Step 3: Ensure secure and responsible AI

From the outset, AI innovation should be delivered with the highest standards for security, assurance, and trust. Beyond taking a leadership role in ensuring safe, secure, and trustworthy AI at a global level, Microsoft provides comprehensive guidance for governments to ensure secure and responsible AI in their efforts, such as the Microsoft Responsible AI Standard, and Microsoft responsible AI practices.   

Continue your modernization journey and become AI ready 

No matter where your government organization stands in its digital transformation, Microsoft and our network of global partners are ready to help you move forward. For more, please explore the following resources: 

• To learn how Microsoft is helping governments solve society’s biggest challenges, see our Microsoft in Government website.  

• For guidance, expert assistance, and resources to help move to Azure, visit Azure Migrate and Modernize and Azure Innovate.  

• To develop a cost management strategy when migrating or operating in the cloud, see Cloud Migration Cost Management Resources.  

¹Microsoft commits $20 billion to advance cybersecurity following meeting with President Biden, Windows Central.

The post Migrate to innovate: How governments are modernizing in advance of AI appeared first on Microsoft Industry Blogs.

The breadth to which Azure OpenAI technologies can be applied is increasing exponentially but must be applied responsibly and in accordance with responsible AI principles and policies.  

As a former defense leader, this blog considers the breadth of opportunities and will highlight three use cases covering the broad spectrum of defense and intelligence missions:

1. Personnel support.
2. Multi-source intelligence analysis.
3. Enterprise knowledge discovery.

These use cases focus on low-classification data, which can be securely optimized by harnessing the collective value and capabilities of Azure OpenAI and the Microsoft Azure Cloud Services. 

Microsoft for defense and intelligence

Power your AI transformation with hyperscale cloud to attain mission success

Explore capabilities

Azure OpenAI use cases    

1. Personnel support. Personnel support spans the spectrum from recruitment to retention and includes numerous capabilities, such as career management, conditions of service, leave, and pay. Azure OpenAI aids this area by providing individuals with answers to their questions across a plethora of personnel-related topics through intuitive chatbots. It also equips decision makers with tools to analyze data, create reports, and make more informed decisions around human resource (HR) policy development. Currently, large quantities of data are stored in enterprise applications or in siloed systems, and substantial levels of resources and time are required to analyze that data and develop the appropriate HR policy proposals. Most notable is the ability to provide valued, timely insights from personnel data at the individual, collective, and enterprise levels as needed. For example, when service personnel are trying to gain policy advice on benefits, Azure OpenAI natural language query capability allows the investigation of policy, with follow-on questions and queries to support an informed decision.   

2. Multi-source intelligence analysis. Multi-source intelligence analysis is a method to gather, process, and interpret information from multiple sources. It involves the integration of data from various intelligence disciplines. Azure OpenAI has the potential to assist analysts as they triage, prioritize, search, analyze, and cross-reference intelligence, ultimately producing actionable information for decision makers at the time of need. Currently, analysts are challenged by an ever-increasing volume, variety, and veracity of data, much of which is unstructured and in different formats. In the future, we envisage Azure OpenAI, cloud-based services, and data being accessible from HQ to the edge. This will allow insights to be derived from both historical and real-time data and deliver actionable intelligence for mission success.  

3. Enterprise knowledge discovery. Defense and intelligence organizations store large quantities of data in enterprise systems that are siloed, and across multiple organizational boundaries. This data is often not clean or structured. Azure OpenAI can expose and correlate this data across different types and sources to find relevant information in response to a natural language query. Examples include querying large repositories of lessons learned from previous operations and exercises matched with current doctrine, and the insights gained from After Action Reviews to support mission planning, simulation, and training for future activities.   

Accessing Azure OpenAI for your organization 

Azure OpenAI has created new possibilities that were once seen as very hard and costly to implement in military systems.  

For ‘non-tactical’ scenarios, cloud-based computing provides the most secure and highest security offering available. The computing resources can be put to work continually enhancing and optimizing planning, analysis, and operational management using the best tools available. Advances in Azure OpenAI and multi-agent frameworks usher in a new era of the role of humans in the loop as a manager and orchestrator of agent computing resources rather than conducting technical analysis and planning. The result is a substantial increase in the speed and capacity of our valuable skilled resources to achieve the mission.  

When we consider ‘tactical’ scenarios, the limitation of bandwidth, weight, and power can influence the adoption of Azure OpenAI applications that can be deployed. Smaller models are less capable and must be finely tuned to their purpose to be highly effective. Additionally, carrying a large number of models not relevant to the mission takes up valuable computing and power resources. As such, when deciding on what Azure OpenAI to access in the field, nations must have robust deployment, collection, and ModelOps pipeline updates that can continually—at speed—update models for specificity and relevance to the tactical edge. The ability to access models in disrupted, disconnected, intermittent, and low-bandwidth (DDIL) environments is essential when operating as close to the edge as Size, Weight, Power, and Compute (SWaP-C) permits. 

Responsible use of AI 

Microsoft is committed to responsible use of AI. That is why Microsoft has long been a leader in ensuring the development of responsible AI, with principles designed to put people first. We believe AI exists to enhance human capabilities, not replace them, and we are committed to empowering responsible AI practices that benefit the world at large. The six key principles for responsible AI include: 

1. Accountability—ensuring transparency and responsibility in AI systems. 
2. Inclusiveness—building AI that considers diverse perspectives and avoids bias. 
3. Reliability and safety—prioritizing safety and robustness in AI deployment. 
4. Fairness—striving for equitable outcomes and avoiding discrimination. 
5. Transparency—providing clear explanations of AI decisions. 
6. Privacy and Security—safeguarding user data and privacy. 

The Microsoft Responsible AI Standard provides actionable guidance for their teams, going beyond high-level principles to create AI systems that uphold these values and earn society’s trust. We also have an Office of Responsible AI that sets governance policies, advises leadership, and ensures responsible practices across the company. 

Begin your AI transformation  

Microsoft Azure OpenAI Service offers unprecedented opportunities to augment human capabilities and enhance decision-making across the defense ecosystem. Harness the value of Azure OpenAI to achieve mission success across the spectrum of capability with greater speed, accuracy, and efficiency.  

To learn more:  

• Attend the AI in the Public Sector Microsoft Build session.
• Visit and bookmark the Microsoft Defense and Intelligence page to stay informed.
• Take the LinkedIn Learning Course, Generative AI for Leaders.
• Reach out to your Microsoft Account Team.

The post Informing defense missions with Microsoft Azure OpenAI Service appeared first on Microsoft Industry Blogs.

It is a remarkable shift, which I and others in Microsoft for government have gauged over the past year as we’ve worked to help city governments solve their most important challenges through technology. 

At the SXSW 2024 conference, the Esri Infrastructure Management & GIS Conference, and in my recent meetings with city leaders from Canada, Finland, and the Netherlands, generative AI has been at the center of most conversations. This excitement is notable because cities are traditionally cautious about technology adoption for important reasons such as risk, privacy, security, and governance, and many are still working through their cloud migration journeys.  

The potential benefits of generative AI to improve operations and service delivery are too compelling for many cities to ignore. To cite just one example, the City of Kelowna in Canada launched an early AI initiative and is using cognitive search and conversational AI to deliver a 24-hour helpline for its 150,000 residents. Project leaders report that generative AI also enables them to automate and streamline internal processes around data entry and analysis, refactor legacy databases and code, and create new apps in minutes.  

This type of innovation has prompted other cities to explore AI innovation, beginning with creating an effective and actionable plan. 

Microsoft for government

Achieve your mission with Microsoft

Explore resources

Success factors for AI adoption in city governments 

Every city is unique, with its own ambitions for the use of generative AI and its own set of requirements and technology considerations. In our work with cities, we have identified a set of success factors that are common across cities investing time and money in generative AI and are enjoying early success.  

1. Empower the workforce with effective upskilling  

Realizing the value of AI starts with the workforce. According to research conducted as part of Microsoft’s Public Sector Insights on Skilling, the lack of skilled workers is often the number one barrier to AI implementation among organizations worldwide.

The imperative to upskill the workforce is particularly important for cities, whose early use cases usually focus on employee productivity and the internal processes they manage. Well-trained and confident workers also help ensure the success of public-facing initiatives. Executive support is key. Workers are empowered when leadership gives them the direction and license to responsibly use AI tools within the context of their day-to-day work. 

To answer the skilling challenge, cities should invest in learning programs, building public-private upskilling partnerships, and giving people adequate time to gain skills and confidence. An ongoing learning experience platform, such as the one developed by Bank of Canada, can promote a culture of learning. Microsoft offers effective resources and strategies, including the Public Sector Center for Digital Skills, which provides specialized insights, guidance, and content, and Microsoft Learn, which offers customized training options. 

2. Build an AI-ready data strategy 

AI is only as good as the data that is made available to it. In city systems, data is often siloed or locked in spreadsheets or other static locations. A modern data strategy is one that integrates such diverse data sources, ensures data quality, establishes rules and processes for data access and management, and keeps data and systems secure.   

An excellent example of how a complete data strategy can deliver ongoing AI benefits is the Smart Qatar (TASMU) Program built by the State of Qatar. Essentially a service platform built on a common data model across multiple domains, TASMU will empower a broad array of AI applications that are expected to help contribute 2% to the nation’s gross domestic product (GDP).  

This sort of comprehensive data strategy is an important long-term goal, but cities should not wait to begin innovation on AI. Many cities are taking an incremental approach, leveraging the quality data they have in hand, with their existing cloud foundations and data governance standards to experiment with new AI use cases. A careful step-by-step approach will guide your data strategy. 

3. Establish frameworks for governance, compliance, and sovereignty  

Some city governments have been reluctant to use AI due to concerns about security, privacy, and compliance requirements. To address these concerns, cities should establish transparent frameworks for AI governance and assurance. The goal is to identify risks and goals associated with both externally facing and internally focused use cases and codify courses of action to ensure success.  

Ensuring the residency of data within strict geographic borders is a key requirement for many cities. This calls for a solution that can ensure the required level of control of sensitive data while still providing a hyperscale cloud environment for a huge number of applications. 

Microsoft offers guidance for cities to establish AI governance and enhance trust and privacy in AI innovation. For cities with strict data residency concerns, we also offer Microsoft Cloud for Sovereignty, which offers tailored cloud services to help build cloud-based workloads in compliance with specific security, policy, and regulatory requirements. 

Take the first steps in building an AI-empowered city  

For city leaders who want to advance their AI journeys, our experts and industry advisors can work with you to identify potential use cases for early innovation based on your specific goals, requirements, and environmental conditions.  

To learn more about how Microsoft can empower cities and government organizations with technology to help solve society’s biggest challenges, visit the Microsoft for government website, read our Microsoft for Government e-book, or get in touch with your Microsoft representative or technology partner.     

The post How Microsoft empowers city governments on the road to AI adoption appeared first on Microsoft Industry Blogs.

During this wave, we’ve invested heavily in the development of copilot templates to enhance capabilities and integration across various industries. These customizable templates offer improved operational efficiency, enhanced customer engagement, and seamless integration with existing technology, all while supporting a diverse, global customer base. Copilots are valuable assets for Microsoft Industry Clouds customers, helping to drive customer and partner success. Microsoft’s partner ecosystem extends our offerings, with systems integrators and independent software vendors enabling factory data ingestion from different systems and building custom UI experiences for the copilot templates on Microsoft Azure AI.

Here’s a look at what’s been delivered since the release plans announcement in January 2024.

Microsoft Cloud for Manufacturing

Microsoft Cloud for Manufacturing is introducing new solutions in preview to optimize factory operations. These include manufacturing data solutions in Microsoft Fabric and a copilot template for factory operations on Azure AI. These solutions enable manufacturers to ingest and unify data from diverse sources, standardize and enrich data for seamless interoperability, and utilize custom copilots for querying data through conversational interfaces. Fabric allows users to maximize the value of factory data and uncover operational insights for production optimization by unifying information and operation technology data into an open and secure data platform. The copilot template for factory operations on Azure AI enhances responsiveness and streamlines communication across teams and roles.

Azure Data Manager for Agriculture

This release of Azure Data Manager for Agriculture includes new copilot templates that can empower organizations to build agriculture copilots with Microsoft Azure OpenAI Service. These templates provide a powerful tool for organizations to use generative AI and data to optimize their operations and engage better with their customers. Customers are bringing generative AI to life for farmers. ITC, a multi-industry enterprise, has created, Krishi Mitra, an AI copilot, developed using Microsoft copilot templates. With this application, ITC seeks to empower farmers by providing them with timely and relevant information that can boost productivity, increase profitability, and enhance climate resilience.

Copilot templates can support use cases based on tillage, planting, crop protection, harvesting, and other types of farm operations. Users can submit queries such as “show me active fields” or “what is the average yield for my field?”. These use cases can help input providers to plan equipment, seeds, applications, and related services and engage better with the farmer.

Using data from Azure Data Manager for Agriculture and other sources, copilots can provide insights on topics like disease risks, yield forecasts, labor needs, crop protection, weather impacts, and harvest windows. Enabling seamless retrieval of data and allowing for plugins, embedded data structures, and subprocesses to be selected as part of the query flow allows organizations to extend their copilot use cases to many roles and scenarios along the agriculture value chain.

Microsoft Cloud for Sovereignty

Microsoft Cloud for Sovereignty is a solution that helps public sector organizations use the public cloud and advanced technologies while helping meet security, sovereignty, and regulatory requirements. The latest release of Microsoft Cloud for Sovereignty introduces updates and new features to streamline the configuration and deployment of sovereign environments. Guardrails and codified architectures reduce complexity and make the process of building sovereign environments more predictable and repeatable. New preview tools include assessment, policy compiler, and drift detection analysis tools, as well as a new Azure service that allow users to create and deploy Sovereign Landing Zones (SLZs) within the Microsoft Azure Portal. Guidance includes sample reference architectures for using large language models (LLMs) and Azure OpenAI Service with SLZ, as well as guidance on workload migrations and Microsoft Power Platform and Microsoft Dataverse configurations.

Microsoft Cloud for Sustainability

In February 2024, Microsoft Cloud for Sustainability announced new data solutions and generative AI advancements in Fabric, providing new levels of speed and efficiency in processing data to help drive faster progress toward sustainability goals. These new features include sustainability data solutions in Fabric and natural language queries with Microsoft Copilot in Microsoft Sustainability Manager, among other AI-powered features now available in preview.

In March 2024, new features were added to Sustainability Manager, including the ability to create calculation models with Copilot using natural language input, a dedicated energy data model to help track energy usage, activity to emissions traceability to link underlying source activity data to emissions records, and the ability to create a Microsoft Power Query template to streamline and accelerate data import.

Additional release updates to Cloud for Sustainability include enhancements to environmental, social and governance (ESG) insights with what-if analysis to help organizations build the relationship between forecasting and reduction goal planning. Users can link forecasts to existing goals to track actual progress alongside the projected ones. In addition, forecasts with the same historical data can be layered onto a single view, allowing for faster analysis of optimal reduction opportunities.

Another new feature is the ability to import and calculate with product carbon footprint data. This feature allows you to use product carbon footprint data to calculate and understand value chain emissions in Sustainability Manager more easily. Organizations can determine the greenhouse gas emissions that are associated with a product family and more easily import and manage this data within Sustainability Manager.

Azure Data Manager for Energy

Azure Data Manager for Energy is expanding geo availability, adding the Australia east region. This additional region is enabled for both the standard and developer tiers of Azure Data Manager for Energy. Users can now select “Australia east” as a preferred region when creating an Azure Data Manager for Energy resource using the Azure portal.

External data sources (EDS) (preview) allow data from external data sources aligned with the OSDU® Technical Standard to be shared with an Azure Data Manager for Energy resource. EDS is designed to pull specified data (metadata) from OSDU-compliant data sources through scheduled jobs while leaving associated dataset files (such as LAS and SEG-Y) stored at the external source for retrieval on demand.

Microsoft Cloud for Retail

Microsoft Cloud for Retail now includes new features in its retail data solutions architecture, an industry-specific workload for unifying, enriching, and modeling industry data on Fabric. Retailers can take advantage of the available list of connectors, application templates, and business intelligence capabilities, which can be easily configured. Retail data solutions offer application templates tailored for retail-specific scenarios, accelerating time to market. These templates serve as customizable and extendable starter kits, allowing retailers to adapt them to their unique requirements. Additionally, application templates and connectors from specialized partners are available. These capabilities enable the seamless use of data to produce unique insights that can’t be achieved in isolation.

One of the new features is the copilot capabilities in Fabric, which bring new ways to transform and analyze data, generate insights, and create visualizations and reports in Fabric and Microsoft Power BI. Another new feature is the Sitecore OrderCloud data connector, which can be used to bring commerce data from Sitecore OrderCloud (preview) into Fabric in real time. The connector performs transformation and orchestration on top of the data from Sitecore OrderCloud to map it to the retail industry data model, reducing engineering effort and accelerating time to insights.

Microsoft Cloud for Healthcare

The 2024 release wave 1 also brings new features and innovations to Microsoft Cloud for Healthcare. One of the new features is the ability to improve clinical and operational insights by ensuring health data is accessible across provider, payor, and pharma; and facilitating clinical, operational, and performance analytics using healthcare data solutions in Fabric (preview).

Some other new features in the 2024 release wave 1 for Microsoft Cloud for Healthcare include support for additional data storage needs, support for availability zones for Microsoft Azure Health Data Services, FHIRLink Power Platform connector, and the ability to use the digital imaging and communications in medicine (DICOM) service with Azure Data Lake integration.

Microsoft Cloud for Nonprofit

Microsoft Tech for Social Impact is proud to announce the April 2024 release for Fundraising and Engagement. This release brings significant enhancements, mainly to nonprofit gift processors, including valuable enhancements to Fundraising and Engagement Azure services and new Stripe API (payment intents) integration. Customers who rely on Stripe for their payment processing can now benefit from the latest Stripe APIs, addressing the requests of current customers and the requirements of future customers. It is highly recommended that customers upgrade and use the new Stripe API when creating a payment processor associated to a configuration profile. For more details, read more here.   

Learn more about Microsoft Industry Clouds

• Official Microsoft Cloud for Manufacturing documentation
• Azure Data Manager for Agriculture Preview
• Microsoft Azure Data Manager for Agriculture Preview
• Official Microsoft Cloud for Sovereignty documentation
• Microsoft Cloud for Sustainability documentation
• Microsoft Azure Data Manager for Energy
• Official Microsoft Cloud for Retail documentation
• Official Microsoft Cloud for Healthcare documentation
• Microsoft Cloud for Nonprofit documentation

Microsoft Industry Clouds

Check out the latest innovations.

Learn more

The post 2024 release wave 1: New copilot features to enhance Microsoft Industry Clouds capabilities appeared first on Microsoft Industry Blogs.

This is of particular interest to our Microsoft for Government team, where our mission is to help governments solve society’s biggest challenges. With the accelerated pace of innovation surrounding generative AI in the past 18 months, we’ve taken a special interest in understanding the full range of factors that either promote or hinder adoption in government organizations. One concern we invariably hear is about skilling—that is, how to provide the training and support required to help employees not just understand the technology but embrace it. Without this, even the best implementations are at risk of failure.

Microsoft for Government

Bridge the AI skills gap in your workforce.

Discover more

How a skilling strategy can advance AI adoption in governments

The growing interest in generative AI has already produced dramatic results in businesses globally, including many noteworthy impacts on governments. Gartner predicts that nearly 25% of governments plan to deploy generative AI solutions by March 2025, with an additional 25% projected the following year.¹ While impressive, this is a bit less than projected global adoption, which is not surprising given that government organizations face an especially high bar in terms of trust, risk, and community expectations.

The promise of generative AI is prompting many governments to complete their cloud migrations and upgrade their data strategies, gradually retiring costly and increasingly inadequate legacy systems. These upgrades will deliver multiple benefits and are essential prerequisites to enabling generative AI innovation.

As it turns out, however, technology alone is not enough. According to a recent IDC study, the number one barrier to implementing and scaling AI is a lack of skilled workers. Nearly 52% of global business leaders named skilling as their top challenge—more than cost (28%), concerns about data or IP loss (28%), or concerns about governance and risk.²

To help fully understand the challenges of skilling that governments face, we commissioned our own research, Public Sector Insights on Skilling. We learned that the obstacles include limited training budgets and a lack of employee time and resources dedicated to learning.

New strategies for upskilling the government workforce for generative AI

Governments who want to be successful with AI should recognize that people are as important as technology in delivering a strong return on investment (ROI). That’s why upskilling is essential, both in terms of funding and employee time.

To help their workforce take advantage of major investments in digital transformation, the Bank of Canada developed an ongoing learning experience platform that provides targeted learning journeys and skills pathways for staff of many specializations. The goal is to free workers from many time-consuming administrative tasks and encourage them to focus on more meaningful work.

Effective practices can be integrated seamlessly into everyday activities, becoming an integral part of the government culture. To help ensure effective upskilling and successful AI adoption, we recommend three important strategies:

1. Make learning an organizational priority

Embracing generative AI is often as much a cultural challenge as a technical one. It starts at the top. Committed, consistent leadership that aligns organizational goals with skilling, and demonstrates learning as a priority, are important factors in creating lasting change. Agencies and organizations need to allocate appropriate resources and understand the gaps in learning, with tactics to fill them. Build a learning culture by allocating time for employee training, identifying opportunities for applying new skills in problem-solving and innovation, and celebrating their success. A long-term mindset is key, as upskilling must be an ongoing employee priority, with opportunities for continuous learning.

2. Build public-private upskilling partnerships

An effective upskilling program is always specific to the government organization. Skilling programs should be developed in the context of their workforces, their local and national requirements, and the unique demands of their agencies and roles. But that doesn’t mean governments need to do it alone. To build a comprehensive approach to embracing AI, you can incorporate upskilling as part of the process from the start. Microsoft, our partners, and other technology providers can help develop skilling strategies. Microsoft offers effective training resources, such as our Public Sector Center for Digital Skills, which offers guidance and resources for public sector organizations in areas such as AI, cloud migration, and cybersecurity.

3. Give people the time to learn

There’s an old Confucian adage that says, “I hear and I forget. I see and I remember. I do and I understand.” Wise words, which in the context of upskilling mean ensuring employees have adequate opportunity to hone their skills. It’s one thing to promise someone that generative AI will save them time. It’s profoundly more impactful when they click on Copilot and generate a document in a fraction of the time it previously required. Such experiences win people over, but they require time. The global head of an IT workplace, interviewed for a Forrester study, estimated that it took employees between four and eight hours of experimenting with Microsoft Copilot for Microsoft 365 before they could use it effectively in their work.³

Empower your government workforce with an effective skilling strategy

We invite you to engage with Microsoft and your local technology partner to develop a skilling strategy as a key part of a holistic approach to AI adoption. For more resources and to learn more about skilling and other training opportunities, please visit:

• Microsoft Learn for government, Microsoft’s flagship resource for technical skilling.
• The AI Skills Initiative, a global resource with free coursework developed in collaboration with LinkedIn.
• AI Explained, a series of virtual community events to educate on generative AI and how it can empower people and organizations (available on-demand in 10 different languages).

Sources:

• ¹ Gartner, “Gartner Predicts Less Than 25% of Government Organizations Will Have Generative AI-Enabled Citizen-Facing Services by 2027”, March 6, 2024.
• ² IDC Infographic, sponsored by Microsoft, The Business Opportunity of AI, IDC #US51315823, November 2023.
• ³ A commissioned study conducted by Forrester Consulting, “New Technology: The Projected Total Economic Impact™ Of Microsoft Copilot For Microsoft 365″, April 2024.

The post 3 ways to solve the skilling challenge for an AI-empowered government workforce appeared first on Microsoft Industry Blogs.

We are excited to announce the latest release of Microsoft Cloud for Sovereignty. This builds upon our December 2023 general availability release, driven by customers and partner feedback, providing tools that simplify the configuration and deployment of complex sovereign controls and expand on best practice guidance. We are also releasing new guidance for comparable approaches to support implementations for Microsoft Power Platform and Microsoft Dataverse. 

Microsoft Cloud for Sovereignty

Meet sovereign requirements with the Microsoft Cloud

Explore capabilities

What’s new in this release

Guardrails—Codified architectures and tooling that reduce complexity and make the process of building sovereign environments designed to help achieve regulatory requirements simple, predictable, and repeatable.  

Tools—New assessment, policy compiler, and drift detection analysis tools to help better manage cloud environments. Introducing a new regional Microsoft Azure service that simplifies the management of Sovereign Landing Zone (SLZ) within the Azure Portal.  

Guidance—Sample reference architectures on how to take advantage of LLMs and Microsoft Azure OpenAI Service based on Retrieval Augmented Generation (RAG) pattern with SLZ and guardrails, as well as guidance on workload migrations. 

Learn more about how to help your organization unlock the power of the hyperscale cloud on the Microsoft Cloud for Sovereignty home and product documentation pages.

Guardrails to simplify building sovereign environments  

We’ve released new features to the Sovereign Landing Zone (SLZ) on GitHub, policy portfolio, and new capabilities of the private preview of Microsoft Cloud for Sovereignty services on the Azure Portal. 

Sovereign Landing Zone on GitHub  

SLZ (generally available) is now configured to use the most recent version of the Azure built-in Sovereignty Baseline Policy Initiative. Similarly, users can also configure the SLZ to deploy any policy set in our portfolio. With this, users will be able to take advantage of our expanding policy portfolio. This capability is now also available for the Azure Landing Zone (ALZ). For more information about when to use either the ALZ or SLZ, review our comparison guidance. 

Users can also configure specific policies (within policy sets) and include additional policies during deployment. For example, customers can easily enable rollout of policies starting with an audit mode and going into an enforcement mode on a granular level. SLZ compliance modules can now be deployed by engineers and administrators to “uplift” an existing landing zone with a similar structure, bringing it closer in line with sovereignty settings and an organization’s requirements. 

Policy portfolio  

Microsoft Cloud for Sovereignty policy portfolio has been updated to better help public sector customers meet several key regulatory frameworks. In collaboration with NATO’s Communication and Information Agency (NCIA), we have reviewed and validated the compliance of cloud deployments with NATO’s D32 directive on information protection (preview). Other updates include custom Azure policy initiatives and control mappings for the Cloud Security Alliance (CSA) Cloud Controls Matrix (now on GitHub), the Netherlands BIO (Baseline Informatiebeveiliging Overheid) Initiative, and the Italian National Cybersecurity Agency custom policy initiatives. These initiatives provide users helpful tools for navigating several diverse regulatory landscapes. 

Tools in preview for Sovereign Landing Zones

Microsoft Cloud for Sovereignty Services on Azure Portal  

Users now have the ability to search for “Microsoft Cloud for Sovereignty” within the Azure Portal. This search will provide visibility to new policy initiatives, documentation regarding the Government Security Program, and transparency logs. Additionally, users can conveniently locate instructions for onboarding into the preview program directly from the portal.  

Managing landing zone configurations in the Azure Portal 

This release streamlines the process of managing landing zone configurations within the Azure Portal. This enables the establishment of efficient and uniform infrastructure at scale by allowing users to create, update, duplicate, generate code, deploy, and delete configurations all from a single pane. For users who wish to customize their deployment, the platform offers the ability to generate a Bicep code package for their infrastructure. Additionally, for organizations seeking a no-code approach to configure and deploy an enterprise-scale landing zone, users can now deploy a landing zone configuration directly from the Azure Portal. 

Sovereignty Baseline Policy Initiatives are automatically assigned to landing zone configurations within management groups, thanks to the fact that Sovereign Baseline Policy Initiatives are now Azure built-in policies. 

Microsoft Cloud for Sovereignty lifecycle tools  

The introduction of new capabilities (preview) offer support for pre-deployment evaluations to streamline policy management processes. These features, available in GitHub, are designed to empower users with greater visibility, control, and compliance across their Azure environments. Our goal is to continually improve the efficiency, reliability, and compliance capabilities of our tools to better meet the evolving needs of our customers. 

Assess and evaluate Azure resources ahead of deployment 

The assessment feature provides a pre-deployment evaluation of Azure resources against established best practices, including the evaluation of resource locations and Azure policy assignments. The tool assesses various aspects, such as the SLZ Baseline Policy assignment, Custom Policy Initiatives usage, and individual policy assignments, offering results categorized as good, better, or best based on severity findings. This is especially helpful during the planning stage of a sovereign implementation and works well with brownfield environments. 

Policy compiler 

The policy compiler is a tool that streamlines the policy management process. It systematically analyzes your organization’s policy initiatives by examining key components—such as display names, descriptions, parameters, and effects. By comparing these elements across different policies, the tool detects redundancies, conflicts, and gaps. It then uses this analysis to provide a set of reconciled policy initiatives, making policy management more efficient and reliable.   

Landing zone drift analyzer 

The drift analyzer monitors and compares the current state of the deployed cloud environment with its original intended landing zone configuration, identifying critical deviations or changes. These deviations, whether intentional or unintentional, are essential indicators of environmental integrity and compliance. Customer feedback will help the evolution of these tools for potential integration with sovereignty services on the Azure Portal. 

Guidance for workload migration and AI configuration

We have recently updated the Microsoft Cloud Adoption Framework to include a sovereignty strategy when using cloud services.  

Our product documents have been updated with guidance on:

• Sovereignty choices for monitoring Azure workloads
• Implement encryption with Customer-Managed Keys in Microsoft Cloud for Sovereignty
• Migrate and modernize with Microsoft Cloud for Sovereignty

This release includes our first guidance and reference architecture around AI and LLM configurations. These articles offer an illustrative example of using LLMs and Azure OpenAI Service within the context of the retrieval augmented generation (RAG) pattern for generative AI. Specifically, it explores how government and public sector organizations can apply these technologies within SLZ while also considering important guardrails.  

We have also recently published guidance on how to configure Microsoft Power Platform and Dataverse environments to improve control over your data and enhance your digital sovereignty posture. This guidance is part of our ongoing efforts to promote digital sovereignty across Microsoft services, including non-Azure services.

Learn more about Microsoft Cloud for Sovereignty

• Microsoft Cloud for Sovereignty 
• Microsoft Cloud for Sovereignty product documentation 
• Sovereign Landing Zone (GitHub) 
• Policy Portfolio (GitHub) 
• Quickstarts (GitHub) 
• Microsoft Cloud for Sovereignty on Azure Portal (preview) 
• Microsoft Cloud for Sovereignty preview (GitHub for preview customers)  
• Log into your Active Azure Directory (AAD) account to request to join Microsoft Cloud for Sovereignty Preview

The post Streamline controls with Microsoft Cloud for Sovereignty appeared first on Microsoft Industry Blogs.