While adoption of AI technology is now widespread, impact is not. Many organizations are experimenting and running pilot programs, but far fewer have the operating discipline to become what we call Frontier Firms—companies that scale AI in ways that meaningfully reshape work, decisions, and value creation. According to IDC’s Business Opportunity of AI Survey (August 2025), 68% of all respondents use GenAI and only 22% of organizations worldwide are Frontier Firms.¹ These companies are seeing a return on investment in the technology that is several times greater than companies that are slow to adopt.

This gap is why Microsoft developed a newly revised 2026 edition of the AI Decision Brief, a handbook designed to help leaders and business decision-makers embrace the opportunities of Frontier Transformation. It addresses how AI can become a durable source of advantage: where to focus, how to measure value, how agents change workflows, and how trust, governance, and responsibility enable scale. “This is not simply the next stage of technology adoption,” writes Brad Smith, Microsoft Vice Chair and President. “Frontier Transformation is a leadership moment that asks organizations to fundamentally rethink how people, processes, and decisions work together.

We believe that this brief answers the questions many executives are asking about how to stay ahead of the curve. The questions below surface what we’re hearing from business leaders across industries as they plan investments, assess readiness, and look ahead. Each reflects a theme explored in depth in the AI Decision Brief and points to how organizations can begin turning AI execution into lasting impact. 

1. How can my company get the biggest impact from AI? 

The biggest impact comes when AI changes how the business operates—not just how fast someone answers an email. “Frontier Transformation is a holistic reimagining of business, aligning AI with human ambition to achieve an organization’s highest aspirations and growth potential,” writes Judson Althoff, CEO of Microsoft commercial business.

What does this mean in practice? Frontier Firms are leveraging AI to transform customer engagement, core processes, decision-making, and innovation. For them, AI isn’t confined to one team or one tool. Instead, it’s embedded across the enterprise in an average of seven business functions. That’s when the outcomes compound. These organizations are monetizing AI and outperforming slow adopters with roughly 3x higher returns.¹ Agents are accelerating that shift because they don’t just make recommendations; they can take action and complete tasks.

2. How do you graduate beyond early wins with AI adoption?

While AI can boost individual productivity—drafting documents, summarizing meetings, and automating the more tedious aspects of jobs—it can do so much more, according to Jaime Teevan, Chief Scientist and Technical Fellow at Microsoft. “The real opportunity is bigger: not just helping individuals work faster, but enabling teams and organizations to work better, together,” she writes. 

Most AI initiatives stall for the same reason most transformations stall: teams prove their value in specific use cases, but leaders don’t change the system around them. The model isn’t the bottleneck—processes, decision rights, and trust are. Frontier leaders, on the other hand, pick a small number of priority workflows and redesign them end to end. That’s how you move from “we got a nice pilot result” to “AI is embedded in how we run the business.”

3. How do I identify the priority workflows where AI can meaningfully change outcomes? 

“AI integration is often framed as a technical problem: which models to use, how to connect systems, how to mitigate risk,” writes Jared Spataro, Microsoft CMO of AI at Work. “But for most organizations, the real constraint on value is not technology, it’s how work is organized and governed. The bigger challenge is centered on management.”

Frontier organizations don’t ask, “Where can we plug in AI to automate a task?” They ask, “Which workflows most directly affect revenue, cost, risk, customer experience, or speed of decision-making?” Frontier leaders focus on embedding AI, agents, and data directly into those areas of high impact. 

4. As AI agents take more action on behalf of employees and teams, how does my role as a leader need to change?

Leadership has become even more important in the agentic era. “When AI systems can plan and execute over many steps, leadership and engineering rigor become the real bottlenecks,” writes Kevin Scott, CTO of Microsoft. “You need teams that are explicit about goals, careful about feedback and evaluation, and thoughtful about where autonomy is earned versus constrained.” 

The greatest risks are unclear intent, ownership, and accountability. Frontier leaders get ahead of this by redefining roles and decision rights early. Humans set outcomes, constraints, and success measures, while agents operate within clearly governed boundaries. That means treating agents like new employees or privileged service accounts—with named owners, least-privilege access, continuous monitoring, and regular review. 

5. How do you measure the success of AI when it’s embedded across workflows, decisions, and teams—not just individual tasks?

“Early productivity gains from AI are now expected,” writes Alysa Taylor, Microsoft CMO of Commercial Cloud and AI. “But Frontier leaders see beyond those short-term efficiency wins. They understand how AI can also help grow revenue, increase customer acquisitions, reshape processes, and improve operational efficiency.” 

Frontier leaders measure ROI the way they run the business: at the workflow and outcome level, not by counting isolated tasks. Yes, they track early productivity signals, but they don’t stop there—they tie AI to business metrics like faster cycle times, higher quality and consistency, better customer experience, lower risk, and faster decision-making.  

6. We’re under pressure to move fast with AI. Can we tackle security later on?

Great question! The answer is simple: absolutely not. “The AI opportunity is incredible, but speed without security, observability and governance opens the door to significant risk. By embedding these elements from the start, organizations can innovate rapidly while building and fostering trust,” writes Vasu Jakkal, CVP of Microsoft Security Business. 

The moment AI moves beyond pilots and starts touching real data, customers, and decisions, issues with security and accountability can offset gains in efficiency. According to Microsoft’s 2026 Data Security Index, less than half (47%) of companies have fully implemented data security controls for AI. Frontier leaders build observability, Zero Trust security, and clear ownership from day one, so teams can move faster with confidence instead of stopping to clean things up later.  

7. How do you scale AI across an organization without losing control or trust?

“Scaling AI is less about deploying tools and more about preparing people,” writes Nathalie D’Hers, Microsoft CVP of Employee Experience. “A workplace culture grounded in a growth mindset is more important than ever.” Frontier Firms embrace continuous learning and agility. This helps teams fundamentally reimagine processes and think bigger.  

Crucially, Frontier organizations also pair empowerment with guardrails. They give employees access to AI where work actually happens—through copilots, low-code tools, and approved platforms—so innovation isn’t bottlenecked by a small group of specialists. At the same time, they’re very clear about boundaries. That includes shared governance frameworks, approved data sources, identity and access controls, and observability at every layer. That’s what allows creation to scale safely.  

8. How do I balance Frontier Transformation with sustainability? 

“AI and sustainability are often treated as separate agenda items, but they are fundamentally connected,” writes Melanie Nakagawa, Chief Sustainability Officer at Microsoft. “Leaders should understand both sides of that equation: the resource footprint of AI as well as the opportunity it brings to help them operate more efficiently, build smarter, more resilient systems, and lower carbon emissions.”  

As AI grows, it brings real resource and trust questions about environmental impact, supply chains, community impact, and whether the benefits of AI are broadly shared. The Frontier view is that designing for efficiency, responsibility, and equitable diffusion isn’t a nice-to-have; it’s how you unlock durable growth while avoiding backlash, constraints, and extra work later.

At Microsoft, we’re building out AI infrastructure with sustainability in mind while also using AI as a force multiplier for climate progress by optimizing systems, accelerating materials discovery, and improving resource efficiency.     

Next steps to lead in the era of Frontier Transformation

Read the full AI Decision Brief to understand what it takes to lead in the era of Frontier Transformation. The insights, leadership advice, and practical tips found within our brief will help prepare your company to properly utilize and scale a powerful AI strategy. Once you have that knowledge base, you’ll need a trusted, reliable set of AI tools to execute that strategy. 

Explore Microsoft AI tools and solutions for your Frontier Transformation. 

¹ IDC InfoBrief: sponsored by Microsoft, What Every Company Can Learn From Frontier Firms Leading the AI Revolution, IDC # US53838325, November 2025 

The post AI Decision Brief: How leaders can drive Frontier Transformation appeared first on The Microsoft Cloud Blog.

This is where the frontier is forming. While copilots help individuals be more effective, agents act on behalf of people. They carry out tasks, orchestrate multi-step workflows, and operate across systems continuously. And they’re moving quickly from experimentation to mainstream use. An IDC InfoBrief, sponsored by Microsoft, shows that 37% of organizations surveyed use agentic AI, another 25% are experimenting with it, and 24% are planning to use it the next 24 months.¹ Organizations that have already invested in people, skills, and responsible practices may be better prepared to operationalize agents at scale—and convert AI’s promise into real business performance.

Five strategic moves for introducing agents responsibly

The new Agents in the Workforce Handbook builds on those earlier foundations. Where the first blog in this series focused on empowering your people, and the second explored how to pair human judgment with AI systems, this third chapter looks ahead: How do you introduce agents into your workforce responsibly and intentionally? Below are five strategic moves leaders should consider. These are high-level guideposts; the Handbook goes much deeper with templates, examples, and decision frameworks to support implementation.

1. Start with your most persistent pain points

When organizations begin exploring agentic AI, a common challenge is prioritization. Imagining use cases is easy. Choosing where to start is harder. Successful organizations don’t begin with futuristic ideas—they begin with the familiar, recurring friction points that quietly drain time and introduce risk.

These are often the workflows teams have learned to “live with”: manual triage, routine follow-up, coordination across systems, repeated reporting steps, or tasks with high error potential. Leaders should observe how work truly happens—shadowing teams, reviewing process maps, and asking simple but revealing questions:

• Where do we lose time?
• What gets done manually that shouldn’t be?
• What feels broken—but no one owns?

These pain points typically offer the clearest path to early value. Addressing them not only frees capacity but also demonstrates to teams how agents can meaningfully improve the day-to-day. The Agents in the Workforce Handbook includes a readiness assessment and real-world patterns to help leaders identify and sequence the right opportunities.

2. Define your AI goal—and lead the change yourself

Introducing agents isn’t only a technical shift—it’s a leadership shift. Frontier Firms choose to align their early agent initiatives around bold, measurable goals: reducing manual work, accelerating cycle times, improving customer responsiveness, or expanding sales capacity. These goals create alignment and momentum, helping teams understand why agents matter and what success looks like.

But goals alone don’t change culture—leaders do. The organizations that move fastest are those whose executives personally model new ways of working. They use agents in their own workflows, talk openly about learnings, and recognize early adopters who demonstrate impact. They also acknowledge that change requires habit‑building. Experimenting with agents for even 20 to 30 minutes a day can materially improve adoption and confidence.

Skilling plays a central role. As Jeana Jorgensen, Corporate Vice President of Global Skilling, notes:

We’re hearing from many of our customers and partners that they expect employees across different roles to spend about 15 to 20% of their week learning and integrating AI into their daily work.

The Handbook offers guidance for identifying the roles, skills, and operating rhythms needed to support agent adoption.

3. Measure what works—and double down where it does

As with any transformative technology, early wins with agents need to be measurable and repeatable. Leaders should ensure visibility into how agents behave, how frequently they’re used, and the outcomes they produce. This isn’t about policing technology—it’s about giving teams the insights needed to improve and scale what’s working.

Effective organizations treat agent adoption like an operational discipline:

• They log and monitor agent activity.
• They measure time saved and business impact generated.
• They expand agents that demonstrate clear value.
• They refine or retire agents that don’t.

These data-driven insights help organizations move from experimentation to a consistent, enterprise-wide model for agent development—one where new ideas become shared services rather than isolated automations. The Handbook goes deeper into measurement strategies, including examples of what high-performing organizations track.

4. As agents become teammates, optimize continuously

Once an organization begins deploying agents across teams, a new challenge emerges: coordination. Agents that start out as individual productivity tools often become shared digital teammates—relied upon by multiple people, processes, and business functions. With that shift comes the need for thoughtful ownership, governance, and communication.

Successful organizations establish clear roles and responsibilities:

• Who owns each agent?
• Who can modify or update it?
• How are changes communicated to the people who rely on it?
• What happens when an agent’s behavior needs tuning?

Agents also require continuous improvement. As they’re used, they encounter edge cases, nuanced team preferences, and shifting processes. Over time, agents become more capable, and employees naturally evolve into “AI managers”—guiding digital apprentices the way they onboard and develop human teammates.

The Handbook provides deeper recommendations for governance models, centers of excellence, and cross-team alignment mechanisms that help organizations scale responsibly.

5. Reinvest the time saved—and push into innovation

While early value often shows up as efficiency, the long-term impact of agentic AI is much bigger: it creates renewed capacity for innovation. Frontier Firms understand that the goal isn’t to simply do the same work faster—it’s to free teams to pursue higher-value ideas, explore new business models, and elevate customer experiences.

Across industries, leading organizations are already demonstrating what this reinvestment looks like:

• L’Oréal is using AI agents to deliver personalized, at-scale beauty guidance.
• AT&T accelerated customer care resolution by 33%, saving millions while improving experience.

These examples highlight a crucial point: agents are not just workflow optimizers. They’re catalysts for reimagining how organizations deliver value. And the companies that begin investing now are positioning themselves for meaningful advantage.

Treat agents like teammates, not tools

The organizations achieving the strongest results view agents not as automations but as digital collaborators—systems that require feedback, tuning, and iteration. They integrate agents into team rhythms, treat them like growing contributors, and help their people evolve into confident AI managers.

This marks the natural third step in the Frontier journey: after empowering employees and strengthening the partnership between human expertise and AI (as explored in the first two blogs), organizations are now ready to bring digital teammates into the workflow in a structured, scalable way.

If your organization is ready to move from experimentation to scaled impact, the Agents in the Workforce Handbook offers the detailed guidance, examples, and templates to support your next phase of Frontier Transformation.

¹ IDC InfoBrief: sponsored by Microsoft, What Every Company Can Learn From Frontier Firms Leading the AI Revolution, IDC # US53838325, November 2025.

The post How to introduce agents into your workforce: 5 actions leaders can take appeared first on The Microsoft Cloud Blog.

Over the past several years, we have seen customer concerns evolve materially. Early conversations focused primarily on privacy and lawful data handling. Today, those concerns have expanded. Leaders are now asking how they maintain operational continuity during disruption, how they adopt AI responsibly without losing control, and how they protect national, organizational, and customer interests in an increasingly volatile global environment.

These questions are not abstract. They surface in boardrooms, procurement decisions, architecture reviews, and crisis simulations. They reflect a broader shift in how trust is evaluated in digital systems. Today in Brussels we brought together attendees from around the world—policy makers, IT leaders, and enterprises—to approach these questions from the multiplicity of perspectives to move the conversation from headlines to action.

From privacy to resilience and beyond

Privacy remains foundational. But it is no longer the sole lens through which sovereignty is assessed.

Customers are increasingly concerned about business continuity in the face of cyber incidents, geopolitical tension, supply chain disruption, and network instability. They want to understand how critical workloads operate if connectivity is constrained, if dependencies fail, or if policy conditions change with little warning.

At the same time, innovation pressures have intensified. AI is becoming central to public service delivery, national competitiveness, and economic growth. Organizations cannot afford to pause progress while sovereignty questions are debated in isolation. They need approaches that allow them to move forward responsibly, balancing opportunity with control.

What we hear consistently is this: sovereignty concerns will continue to evolve. Any approach that treats them as static is already behind.

For four decades, Microsoft has operated under some of the world’s most demanding data protection, competition, and digital governance frameworks. Working closely with European institutions, regulators, and customers has shaped how we think about sovereignty—not as a regional exception, but as a discipline that must function at scale, under scrutiny, and over time. That experience matters because many of the sovereignty questions now emerging globally were first tested in Europe, long before they became mainstream elsewhere.

A consultative approach to risk management

This is why we believe digital sovereignty must be approached as consultative risk management, not a checkbox or a predefined deployment model.

Every organization faces a unique mix of regulatory obligations, cyber risk, operational exposure, and innovation goals. Even within a single institution, sovereignty requirements differ by workload. Some demand strict isolation and local control. Others require global scale, advanced security capabilities, and rapid innovation.

Our role is to help customers navigate these tradeoffs deliberately. That means working with them to assess risk, align architecture to policy realities, and design environments that reflect both today’s constraints and tomorrow’s unknowns.

This work sits at the intersection of cybersecurity, compliance, resilience, and frontier transformation. It requires ongoing engagement, transparency, and the willingness to adapt as conditions change.

Digital sovereignty posture in practice

A digital sovereignty posture that is flexible recognizes that no single approach can address every requirement. Instead, it focuses on giving organizations options, visibility, and control across a continuum of environments.

Customers operating in public cloud environments expect clear data residency options, strong encryption and access controls, and visible operational discipline. Just as important, they look for transparency into how cloud systems are governed and how exceptional situations are managed, particularly as regulatory scrutiny increases.

Those expectations do not disappear when workloads move closer to the edge. In fact, they intensify. For workloads that require greater isolation, local processing, or operation in constrained environments, hybrid and disconnected solutions become essential. In February, Microsoft announced the expansion of disconnected operations, enabling customers to run critical workloads in air-gapped environments while retaining consistent governance and operational control. This capability extends cloud-based practices into disconnected settings, supporting operational continuity without abandoning security and innovation. 

That commitment shows up in concrete safeguards that customers can independently evaluate and apply. The EU Data Boundary is one example, supporting data storage and processing within the EU and European Free Trade Association (EFTA) regions for cloud services, alongside longstanding investments in encryption, access controls, auditability, and operational transparency. These measures provide practical mechanisms for aligning cloud operations with regulatory and risk requirements, rather than relying on abstract assurances. 

At the same time, we are expanding options across hybrid and private cloud environments to support continuity, resilience, and local control where required. These investments reflect a simple reality: customer needs are not converging toward one model. They are diversifying.

Underpinning all of this are Microsoft’s digital commitments, which frame how we approach privacy, security, transparency, and responsible AI. These commitments are not marketing statements. They guide how systems are built, operated, and governed, and they provide a foundation for long-term accountability.

Practical guidance for leaders navigating sovereignty

As digital sovereignty becomes embedded in policy and procurement decisions, leaders benefit from a practical lens. Based on what we hear from customers and stakeholders, there are a few consistent themes shaping successful approaches:

• Sovereignty requirements will continue to expand beyond privacy to include continuity, resilience, and AI governance.
• Risk management is now inseparable from digital transformation strategy.
• Flexibility and optionality matter more than rigid architectures.
• Transparency and accountability are as important as technical capability.
• Sovereignty posture must consider protections against cyberthreats.

Addressing these realities requires partners who understand the full scope of the challenge and are willing to engage over the long term. It requires platforms and collaboration designed with sovereignty in mind from the start.

So what does this mean for you?

Digital sovereignty is not a destination. It is an ongoing discipline shaped by changing technology, regulation, and global conditions.

At Microsoft, we approach this work with humility and responsibility. We recognize that customer concerns will continue to evolve, and that our own platforms and practices must evolve with them. We remain committed to expanding our sovereign cloud continuum, strengthening our cloud capabilities, and delivering solutions that balance innovation with control.

Most importantly, we remain focused on delivery. Because in moments of uncertainty, what matters most is not what technology promises, but what it allows organizations to do with confidence.

Where does digital sovereignty go from here?

The future of digital sovereignty will be defined by implementation, not rhetoric. Success will depend on collaboration between governments, industry, and civil society, as well as a shared commitment to transparency and continuous improvement.

As we look ahead, our focus remains on helping organizations turn sovereignty principles into durable, scalable outcomes. That means continuing to invest in capabilities that support trust, engaging constructively with policymakers, and listening closely to the evolving needs of our customers.

Digital trust is built over time, through consistent action and openness, and that trust is one of the most important foundations we can help create.

The post Navigating digital sovereignty at the frontier of transformation appeared first on The Microsoft Cloud Blog.

Deepfakes threaten trust in news, elections, brands and everyday interactions, leading us to question what’s real. Determining what’s authentic or manipulated is the subject of Microsoft’s “Media Integrity and Authentication: Status, Directions, and Futures” report, published today. The study evaluates today’s authentication methods to better understand their limitations, explore potential ways to strengthen them and help people make informed decisions about the online content they consume.

The authors conclude that no single solution can prevent digital deception on its own. Methods such as provenance, watermarking and digital fingerprinting can offer useful information like who created the content, what tools were used and whether it has been altered.

People can be deceived by media if they lack information like its origin and history, or if its information is low-quality or misleading. The goal of the report is to provide a roadmap to deliver more high-assurance provenance information the public can rely on, according to Jessica Young, director of science and technology policy in the Office of the Chief Scientific Officer at Microsoft.

Helping people recognize higher-quality content indicators is increasingly important as deepfakes become more disruptive and provenance legislation in various countries, including the U.S., introduce even more ways to help people authenticate content later this year.

Media provenance has been evolving for years, with Microsoft pioneering the technology in 2019 and cofounding the Coalition for Content Provenance and Authenticity (C2PA) in 2021 to standardize media authenticity.

Young, co-chair of the study, explains more about what it all means:

What prompted the study?

“The motivation was two-fold,” Young says. “The first is the recognition of the moment we’re in right now. We know generative AI capabilities are becoming increasingly powerful. It’s becoming more challenging to distinguish between authentic content — like content that was captured by a camera versus sophisticated deepfakes — and as a result, there’s a huge uptick right now in interests and requirements to use those technologies that exist to disclose and verify if content was generated or manipulated by AI.

“The moment has been building, and we have a desire to help ensure that these technologies ultimately drive more benefit than harm, based on how they’re used and understood.”

Young adds that the paper is meant to inform the greater media integrity and authentication ecosystem, including creators, technologists, policymakers and others to understand what is and isn’t possible currently and how we can build on it for the future.

What did the study accomplish, and what did you learn?

The report outlines a path to increase confidence in the authenticity of media. The authors propose a direction they refer to as “high-confidence authentication” to mitigate the weaknesses of various media integrity methods.

Linking C2PA provenance to an imperceptible watermark can bring relatively high confidence about media’s provenance, she says.

She notes the report has a lot of caveats too, such as how provenance from traditional offline devices like cameras, which often lack critical security features, can be less trustworthy because it’s easier to alter.

It isn’t possible to prevent every attack or stop certain platforms from stripping provenance signals, so the challenge, Young says, “is figuring out how to surface the most reliable indicators with strong security built in — and, when necessary, reinforce them with additional methods that allow recovery or support manual digital-forensics work.”

How is this study different from others?

Young says their study investigated two “underexplored” lines of thought for the three methods of verification. They define the first as sociotechnical attacks, where provenance information or the media itself could be manipulated to make authentic content appear synthetic or fake content seem real during the validation process.

“Imagine you see an authentic image of a global sporting event with 80% of the crowd cheering for the home team,” she says. “The away team engages in an online argument claiming, ‘Hey, no, that’s all a fake crowd.’ Someone could make one small, insignificant edit to a person in the corner of the picture and current methods would deem it AI generated — even if the crowd size was real. These methods that are supposed to support authenticity are now reinforcing a fake narrative, instead of the real one.

“So, knowing how different validators work, even through really subtle modifications, you could manipulate the results the public would see to try to deceive them about content,” she says. The second key topic builds on the C2PA’s work to make content credentials more durable, while also addressing reliability. This is where the research is especially novel, Young says. “We looked at how provenance information can be added and maintained across different environments — from high-security systems to less secure, offline devices — and what that means for reliability.”

Why is verifying digital media so difficult?

Authenticating media is complex because there’s not a one-size-fits-all solution, Young says.

“You have different formats that have different limitations or trade-offs for the signals they can contain,” she explains. “Whether it’s images, audio, video — not to mention text, which has a whole different array of challenges — and how strong the solutions can be applied there.”

Young says there are different requirements and opinions about what level of transparency is appropriate as well. In some cases, users might not want any of their personal information included in the digital provenance of a piece of media, while in others, creators or artists might want attribution and to opt-in for having their information included.

“So, you have different requirements or even considerations about what goes into that provenance information,” she says. “And then, similar to the field of security, no solution is foolproof. So, all the methods are complementary, but each has inherent limitations.”

Where do we go from here?

Young says that as AI-made or edited content becomes more commonplace, the use of secure provenance of authentic content is becoming increasingly important. Publishers, public figures, governments and businesses have good reason to certify the authenticity of the content they share. If a news outlet shoots photos of an event, for example, tying secure provenance information to those images can help show their audience the content is reliable.

“Government bodies also have an interest in the public knowing that their formal documents or media are reliable information about public interest matters,” Young says.

She adds that as AI modifications to media become “increasingly common” for legitimate purposes, secure provenance can provide important context to help prevent an average reader or viewer from simply dismissing that content as fake or deceptive.

“For the industry and for regulators, we note how important continued user research in this area is to drive towards more consistent and helpful display of this information to the public — to make sure it’s actually meaningful and useful in practice,” Young says.

“We have a limited set of technologies that can assist us, and we don’t want them to backfire from being misunderstood or improperly used.”

Learn more on the Microsoft Research Blog.

The post A new study explores how AI shapes what you can trust online appeared first on The Microsoft Cloud Blog.

Microsoft Global Skilling helps people and organizations build the skills they need to thrive in an AI‑powered world. Within Global Skilling, the Learning Lab is the innovation engine—a team focused on designing, testing, and evolving modern learning experiences to continuously improve how skills are developed, validated, and applied in the flow of work. 

AI is reshaping how organizations work. Teams aren’t just adopting new tools—they’re also figuring out how those tools fit into existing workflows, roles, and expectations, all while trying to keep pace with business demands in a rapidly changing landscape. It’s a heavy lift. As the leader of the Learning Lab team, I’m navigating these same pressures, along with my team members, as we balance day-to-day delivery with the need to evolve our processes in real time. That’s why we’re embedding AI assistants and agentic workflows into internal processes—using them not only to work differently but also to learn differently. Through experimentation, we’re uncovering new ways to streamline operations and improve the learner experience for our global audience.  

This blog highlights three of our team’s most impactful AI initiatives that could also benefit your organization. Inspired by these projects, we developed A Practical Guide for Bringing AI into Your Business Processes, featuring real-world examples and actionable ideas for integrating AI and human expertise across your organization. 

A Practical Guide for Bringing AI into Your Business Processes

Download the guide

3 impactful AI initiatives leading the way

1. Reducing time-intensive coordination to optimize research 

The challenge of coordinating teams for research  

Before any learning materials can be built, our team conducts extensive research to understand new technologies, identify required skills, and validate what learners need. This early-stage analysis requires input from multiple stakeholders and a deep review of internal documentation, product roadmaps, and existing training materials.  

How AI is helping accelerate our research tasks and optimize cross-team input 

One of the biggest bottlenecks for our research workflows has been the time it takes to synthesize information and align teams around what a course should achieve. To improve this, we began experimenting with Researcher in Microsoft 365 Copilot and persona-based agents to support our research and planning stages. Our new process looks like this: 

• Researcher synthesizes internal documentation, product roadmaps, and existing training materials to surface emerging themes and identify knowledge gaps. With the ability to process thousands of pages in minutes, it flags potential course objectives the team might have missed.
• In parallel, persona-based agents simulate the perspectives of stakeholders from varying teams to help validate ideas before bringing them to the key decision-makers.
• Throughout this process, our team members guide these AI tools through every step—providing the business context, analyzing AI outputs to identify gaps or inconsistencies, refining direction, and ensuring consideration of broader business objectives.  

In our experience with AI handling synthesis and early-stage validation, we’ve reduced the time required for core research processes from two weeks to just one day. This significant time savings extends to every course developed with this method, enabling us to redirect focus toward shaping stronger strategies, aligning content with business impact, and accelerating decision-making across teams.

Applying this approach in your organization 

AI-supported research and planning can help you make sense of complex information faster and build alignment earlier in your decision cycles. By using AI to synthesize documents, surface patterns, and validate assumptions, you can reduce the effort required to get teams on the same page. Your team members can then focus on refining strategy, confirming business priorities, and shaping higher-impact decisions. This combination improves speed and clarity throughout cross-functional work.  

Explore A Practical Guide for Bringing AI into Your Business Processes to learn more about how you can apply this in processes like: 

• Drafting onboarding plans that human resources (HR) leaders can tailor to company culture.
• Developing quarterly sales plays informed by shifting buyer behavior and competitor activity.
• Creating campaign briefs rooted in audience insights, market trends, and performance data.
• Developing forecasting assumptions by synthesizing inputs from sales, operations, and historical data. 

2. Transitioning from manual maintenance to continuous quality improvements 

The challenge of shorter content lifecycles  

We maintain thousands of courses and lab environments as part of our skilling initiatives for Microsoft technologies. With the fast pace of product evolution, it can be challenging to keep learning content accurate and functional.  

How GitHub Copilot became the maintenance partner for the team 

We recognized that the demands for maintaining learning content were increasing beyond our capacity to manage effectively. So we integrated GitHub Copilot into the content maintenance workflow like this: 

• GitHub Copilot tools analyze content repositories—flagging inconsistencies, identifying outdated examples, and recommending updates based on current documentation.
• Throughout this process, our team reviews and refines the AI-generated recommendations. When GitHub Copilot flags an issue, we evaluate how those changes might apply to other training courses. We also ensure that all revisions align with learning objectives and verify that security and accessibility standards are met.
• Then GitHub Copilot helps implement some of the suggested updates, like generating new code samples or suggesting environmental configurations that align with the latest product releases. 

As a result, our team has reduced the time we spend on routine content maintenance by up to 25%. And with these time savings, team members can shift from reactive updates to proactive innovation—evaluating emerging skills, shaping next-generation modules, and exploring how agents, simulations, and personalized learning could improve outcomes. 

Applying this approach in your organization 

AI-assisted maintenance can help you keep large, fast-changing content ecosystems accurate and up to date without overwhelming your teams. By using AI to surface inconsistencies, flag outdated material, and recommend updates, you can dramatically reduce time spent on routine fixes. Your experts can then focus on reviewing changes for accuracy, regulatory needs, and strategic intent. This balance enables you to maintain quality at scale while freeing your teams to invest in higher-value innovation.  

Explore A Practical Guide for Bringing AI into Your Business Processes to learn more about how you can apply this in processes like: 

• Maintaining and updating sales enablement content as product and service offerings evolve.
• Keeping product messaging frameworks and campaign assets consistent and up to date.
• Updating help center articles and support workflows after feature releases.
• Updating contract templates and clause libraries to align with new regulatory guidance.

3. Delivering inclusive learning at scale through diverse content formats 

The challenge of content relevance and engagement  

Our learners span every continent, speak dozens of languages, and have their own preferred learning methods. Creating multimodal, accessible, and inclusive learning experiences while managing constant content updates was stretching the team thin.  

How AI helps scale and translate content for global learners  

To support different learning styles and languages, we’re piloting how to create immersive, inclusive learning through two experiments with AI: 

1. We’re using AI tools to turn a single source of training content, like a session transcript or recording, into multiple formats, such as videos, podcasts, and recap summaries. This multimodal output lets us update learning materials at the pace required by our global audience and helps ensure that we’re reaching learners in their preferred formats.
2. We’re piloting an AI-powered tool that not only translates content but also generates avatars that deliver multilingual voiceovers with more natural lip-sync, eliminating one of the most distracting elements of dubbed content. 

Early results show that we can now recover up to 15 hours per course we develop—time our team can spend on more nuanced work that AI can’t do, like adapting cultural references, verifying that tone and pacing match learning objectives, and maintaining brand voice. 

Applying this approach in your organization 

AI-powered localization can help you deliver content that feels native to every audience you service, no matter the language or market. By pairing AI’s speed in translation, voiceover, and prompt generation with your team’s expertise in cultural nuance and brand standards, you can scale global engagement without diluting quality. This combination lets you reach more learners, customers, and employees while keeping your message consistent and relevant across regions.  

Explore A Practical Guide for Bringing AI into Your Business Processes to learn more about how you can apply this in processes like: 

• Localizing campaign assets for regional markets across languages and cultural norms.
• Tailoring pitch decks and demos for industry-specific or region-specific buyers.
• Creating multilingual chatbot responses and support scripts for global customers.
• Adapting standard operating procedure and process documentation for different facilities or regional regulations. 

Building skills and strengthening our AI strategy

As AI becomes an extension to the Learning Lab, we’ve discovered that it’s much more than just implementing new tools—it’s also a journey of building technical and human skills across the team. Our experiments require every team member to stretch into new capabilities, from process optimization and innovation to strengthening collaboration and creative problem-solving. As a result, we’ve been able to spend less time on repetitive tasks and to dedicate more energy to the kind of creative, relationship-driven work that leads to exceptional learning experiences. 

Looking to build skills for you and your teams? Explore AI Skills Navigator, the agentic learning space that brings together AI-powered skilling experiences and credentials that help individuals build career skills and organizations worldwide accelerate their business.

The post How to bring human expertise and AI together: 3 impactful initiatives appeared first on The Microsoft Cloud Blog.

Read the latest Cyber Pulse report

Agent building isn’t limited to technical roles; today, employees in various positions create and use agents in daily work. More than 80% of Fortune 500 companies today use AI active agents built with low-code/no-code tools.² AI is ubiquitous in many operations, and generative AI-powered agents are embedded in workflows across sales, finance, security, customer service, and product innovation. 

With agent use expanding and transformation opportunities multiplying, now is the time to get foundational controls in place. AI agents should be held to the same standards as employees or service accounts. That means applying long‑standing Zero Trust security principles consistently:

• Least privilege access: Give every user, AI agent, or system only what they need—no more.
• Explicit verification: Always confirm who or what is requesting access using identity, device health, location, risk level.
• Assume compromise can occur: Design systems expecting that cyberattackers will get inside.

These principles are not new, and many security teams have implemented Zero Trust principles in their organization. What’s new is their application to non‑human users operating at scale and speed. Organizations that embed these controls within their deployment of AI agents from the beginning will be able to move faster, building trust in AI.

The rise of human-led AI agents

The growth of AI agents expands across many regions around the world from the Americas to Europe, Middle East, and Africa (EMEA), and Asia.

According to Cyber Pulse, leading industries such as software and technology (16%), manufacturing (13%), financial institutions (11%), and retail (9%) are using agents to support increasingly complex tasks—drafting proposals, analyzing financial data, triaging security alerts, automating repetitive processes, and surfacing insights at machine speed.³ These agents can operate in assistive modes, responding to user prompts, or autonomously, executing tasks with minimal human intervention.

And unlike traditional software, agents are dynamic. They act. They decide. They access data. And increasingly, they interact with other agents.

That changes the risk profile fundamentally.

The blind spot: Agent growth without observability, governance, and security

Despite the rapid adoption of AI agents, many organizations struggle to answer some basic questions:

• How many agents are running across the enterprise?
• Who owns them?
• What data do they touch?
• Which agents are sanctioned—and which are not?

This is not a hypothetical concern. Shadow IT has existed for decades, but shadow AI introduces new dimensions of risk. Agents can inherit permissions, access sensitive information, and generate outputs at scale—sometimes outside the visibility of IT and security teams. Bad actors might exploit agents’ access and privileges, turning them into unintended double agents. Like human employees, an agent with too much access—or the wrong instructions—can become a vulnerability. When leaders lack observability in their AI ecosystem, risk accumulates silently.

According to the Cyber Pulse report, already 29% of employees have turned to unsanctioned AI agents for work tasks.⁴ This disparity is noteworthy, as it indicates that numerous organizations are deploying AI capabilities and agents prior to establishing appropriate controls for access management, data protection, compliance, and accountability. In regulated sectors such as financial services, healthcare, and the public sector, this gap can have particularly significant consequences.

Why observability comes first

You can’t protect what you can’t see, and you can’t manage what you don’t understand. Observability is having a control plane across all layers of the organization (IT, security, developers, and AI teams) to understand:  

• What agents exist 
• Who owns them 
• What systems and data they touch 
• How they behave 

In the Cyber Pulse report, we outline five core capabilities that organizations need to establish for true observability and governance of AI agents:

• Registry: A centralized registry acts as a single source of truth for all agents across the organization—sanctioned, third‑party, and emerging shadow agents. This inventory helps prevent agent sprawl, enables accountability, and supports discovery while allowing unsanctioned agents to be restricted or quarantined when necessary.
• Access control: Each agent is governed using the same identity‑ and policy‑driven access controls applied to human users and applications. Least‑privilege permissions, enforced consistently, help ensure agents can access only the data, systems, and workflows required to fulfill their purpose—no more, no less.
• Visualization: Real‑time dashboards and telemetry provide insight into how agents interact with people, data, and systems. Leaders can see where agents are operating, understanding dependencies, and monitoring behavior and impact—supporting faster detection of misuse, drift, or emerging risk.
• Interoperability: Agents operate across Microsoft platforms, open‑source frameworks, and third‑party ecosystems under a consistent governance model. This interoperability allows agents to collaborate with people and other agents across workflows while remaining managed within the same enterprise controls.
• Security: Built‑in protections safeguard agents from internal misuse and external cyberthreats. Security signals, policy enforcement, and integrated tooling help organizations detect compromised or misaligned agents early and respond quickly—before issues escalate into business, regulatory, or reputational harm.

Governance and security are not the same—and both matter

One important clarification emerging from Cyber Pulse is this: governance and security are related, but not interchangeable.

• Governance defines ownership, accountability, policy, and oversight.
• Security enforces controls, protects access, and detects cyberthreats.

Both are required. And neither can succeed in isolation.

AI governance cannot live solely within IT, and AI security cannot be delegated only to chief information security officers (CISOs). This is a cross functional responsibility, spanning legal, compliance, human resources, data science, business leadership, and the board.

When AI risk is treated as a core enterprise risk—alongside financial, operational, and regulatory risk—organizations are better positioned to move quickly and safely.

Strong security and governance do more than reduce risk—they enable transparency. And transparency is fast becoming a competitive advantage.

From risk management to competitive advantage

This is an exciting time for leading Frontier Firms. Many organizations are already using this moment to modernize governance, reduce overshared data, and establish security controls that allow safe use. They are proving that security and innovation are not opposing forces; they are reinforcing ones. Security is a catalyst for innovation.

According to the Cyber Pulse report, the leaders who act now will mitigate risk, unlock faster innovation, protect customer trust, and build resilience into the very fabric of their AI-powered enterprises. The future belongs to organizations that innovate at machine speed and observe, govern and secure with the same precision. If we get this right, and I know we will, AI becomes more than a breakthrough in technology—it becomes a breakthrough in human ambition.

Get the full Cyber Pulse report

To learn more about Microsoft Security solutions, visit our website. Bookmark the Security blog to keep up with our expert coverage on security matters. Also, follow us on LinkedIn (Microsoft Security) and X (@MSFTSecurity) for the latest news and updates on cybersecurity.

¹Microsoft Data Security Index 2026: Unifying Data Protection and AI Innovation, Microsoft Security, 2026.

²Based on Microsoft first‑party telemetry measuring agents built with Microsoft Copilot Studio or Microsoft Agent Builder that were in use during the last 28 days of November 2025.

³Industry and Regional Agent Metrics were created using Microsoft first‑party telemetry measuring agents built with Microsoft Copilot Studio or Microsoft Agent Builder that were in use during the last 28 days of November 2025.

⁴July 2025 multi-national survey of more than 1,700 data security professionals commissioned by Microsoft from Hypothesis Group.

Methodology:

Industry and Regional Agent Metrics were created using Microsoft first‑party telemetry measuring agents built with Microsoft Copilot Studio or Microsoft Agent Builder that were in use during the past 28 days of November 2025. 

2026 Data Security Index: 

A 25-minute multinational online survey was conducted from July 16 to August 11, 2025, among 1,725 data security leaders. 

Questions centered around the data security landscape, data security incidents, securing employee use of generative AI, and the use of generative AI in data security programs to highlight comparisons to 2024. 

One-hour in-depth interviews were conducted with 10 data security leaders in the United States and United Kingdom to garner stories about how they are approaching data security in their organizations. 

Definitions: 

Active Agents are 1) deployed to production and 2) have some “real activity” associated with them in the past 28 days.  

“Real activity” is defined as 1+ engagement with a user (assistive agents) OR 1+ autonomous runs (autonomous agents).  

The post 80% of Fortune 500 use active AI Agents: Observability, governance, and security shape the new frontier appeared first on The Microsoft Cloud Blog.

In this Q&A, Michelle introduces M12, considers what kinds of AI-powered solutions will drive the next wave of AI innovation, highlights some of M12’s recent investments, and explains how Microsoft Marketplace can help startups reach enterprise customers.

What is M12, and how does it differ from traditional venture capital firms? 

M12 is Microsoft’s venture fund. We’re an early-stage fund focused on finding innovative AI startups and technologies that can leverage the power of Microsoft. And with that, we’re looking at the big picture—not just the companies that are winning, but those that are driving the systems of the future.

In today’s venture environment, there’s a lot of capital fighting to invest in the top AI founders and startups, so investors need something extra to stand out and win deals.

At M12, we’re backed by the power of Microsoft, which means a few things: instant credibility with potential partners, as well as access to its ecosystem,  expertise and research, and global go-to-market (GTM) motion. But most importantly, the team is built for impact. We help ensure that our startups have a custom plan and a dedicated resource to help make that plan a reality.  

Our team is a unique mix of founders, researchers, and folks who have been in the trenches across product innovation, finance, go to market, and business development. While our skill sets may be diverse, we put our collective expertise toward a common goal: to help promising companies win.  

What kinds of AI-powered solutions or business models do you believe will define the next wave of innovation? 

The last few years we’ve seen a lot of experimentation—thousands of AI fueled products have been launched—some finding spectacular product market fit, reaching more than USD100 million annual recurring revenue (ARR) in less than a year of launch, while others stuck in the pilot stage with low adoption or questionable durability.

We believe we are now entering a new phase of AI that’s less about experimentation and more about putting into production and delivering measurable outcomes. Buyers are holding new technology accountable to real return on investment (ROI) on shorter time horizons—is your product saving the enterprise money, driving measurable efficiencies, or generating new revenue opportunities.

We see the next wave of application technology embedded deeper into enterprise workflows, training on proprietary data, more domain-aware agents that move beyond assistance to coordinating multi-step work across teams, and orchestration tools that go beyond copilots. We believe that early use cases of AI like software coding and customer support will continue to scale, but it’s been interesting to see momentum in fields that traditionally have been slower to adopt new technology such as medicine, law, and accounting.

At M12, we’ve been thinking deeply about the next evolution in models beyond text-based large language models (LLMs), particularly world models, and how data from the physical world brings about new innovations, particularly in science.

We are also focused on the foundational layers that make AI possible at scale— infrastructure, tooling, data systems, and how to make AI factories more efficient, sustainable, and performant. You can see that in our recent investments in nEye.AI and Neurophos, and you’ll continue to see us make bets in these areas.

On the business side, we’re seeing rapid change. New pricing models like outcome-based and consumption-based pricing, faster adoption cycles particularly in small and midsize business (SMB) and individual buyers, and a sharper focus on ROI are reshaping how value is created.

Achieve business goals with AI

Explore extensive guidance and resources from Microsoft

Get access

What are the key criteria you look for when evaluating businesses to invest in?

We look at a combination of factors and because we invest early in a company’s lifecycle, we focus deeply on the strength of the team, their conviction, and their long-term vision. The best teams are shipping quickly, adopting the latest models, listening closely to customers, and adjusting based on real-world feedback. We are finding that velocity is “the” moat for many AI companies. We also look at market size and opportunity and increasingly, a founder’s plan for how to differentiate in a crowded market, and how we can support the business.  

Ultimately, we’re looking for incredible teams that are learning quickly, innovating with customers, and building with durability in mind. 

What are some of your recent investments and what made them stand out?

M12 meets hundreds of startups each year and typically invests in 18 to 20 early-stage companies a year. We aim to balance our portfolio with a mix of companies that have nearer term adoption and traction and those with longer time horizons.

• Inception Labs is pioneering a diffusion large language model (LLM) approach, which has some performance advantages over transformers. Its early—a “foundation model” and frontier investment that we’re excited by.
• Outset is an AI-moderated research platform that is fundamentally reshaping how enterprises operate and understand their customers, which is exactly what we look for in teams and technologies—they already have an impressive list of customers, including Microsoft, Uber, and Hubspot. 
• Neurophos is tackling one of the biggest bottlenecks in AI infrastructure: the inability of GPUs to scale efficiently across cost, power, and footprint. The team has developed a manufacturable optical processing unit that compresses GPU-scale compute into a dramatically smaller, cooler, and more sustainable form, delivering up to 100× performance and energy efficiency compared to today’s leading systems.
• Entire is approaching the next general wave of developer platforms from first principles, architecting a new platform from the ground up for agent-to-human collaboration. We’re excited to invest in Entire and support a team that is inventing a new paradigm of how developers and AI work together.

Where do you see the most promising opportunities for AI to disrupt industries globally, and how is M12 positioning itself to lead in those areas?

The biggest opportunities for AI disruption are in industries where complexity, scale, and constraints intersect—areas like enterprise transformation (this could be system of records being unbundled or completely rethought, hyper-personalized employee agents and applications), next gen cybersecurity, tooling and infrastructure that needs to be rebuilt when agents are writing the majority of software code, and innovations brought on by world models deployment are also top of mind.

M12 is uniquely positioned to support founders tackling these challenges because we invest where Microsoft brings deep expertise—enterprise environments, technology, global scale, and mission-critical systems. Our investors are thesis driven and often will spend months getting to know a space, going deep, which builds credibility with founders before an investment.  

How are you seeing founders adapt their strategies to thrive in an AI-first world, and what advice would you give them?

Founders are shipping and executing faster, focusing on deeper integrations with customers, and building community and brand awareness quickly as part of their go-to-market. Many enterprises today are running multiple proofs of concept across similar AI tools and we believe this next year customers will focus and consolidate. There will be a major shift towards deciding which solutions deliver real ROI and are ready for wide-scale production. 

Founders who succeed will be the ones who build deeply integrated, customized workflows that fit into how customers actually operate. We’re seeing startups pairing product innovation with services, like forward-deployed engineers, to accelerate adoption early on, getting feedback loops started quickly as well with access to unique data to make their products sticky and “must haves.”

We’re also seeing companies rethink how they operate internally. One of our portfolio companies, Allstacks, which was founded more than 5 years ago, radically shifted their go-to market strategy last year to become AI native, where AI tools and agents are integral parts of each operating function. It was critical work and a good example of how forward-thinking companies are reinventing their operations to match their product innovation.

Many AI native companies are operating incredibly leanly, even as they scale, due to the use of AI in all elements and functions of their business. 

How does M12 evaluate which AI startups have the potential to scale and make a lasting impact?

We spend a lot of time distinguishing between the hype cycle and the durability cycle. Markets can move quickly, we’re seeing that more than ever, but lasting companies are built deliberately over many years.

The startups with the greatest potential are those that design with constraints in mind, whether that’s power, capital efficiency, inference economics, or customer trust, and still find ways to deliver meaningful value.

Why is collaboration between corporations and startups essential for accelerating AI transformation globally?

AI transformation and adoption especially in the enterprise doesn’t happen in isolation. Startups bring speed, creativity, and new ideas. Corporations bring scale, trust, and real-world deployment environments.

When those strengths come together, we see not an additive but multiplier effect. Corporations like Microsoft can serve as a bridge between early-stage founders and global enterprise customers, helping startups move faster while deploying AI responsibly and at scale.  

For example, Microsoft Marketplace is a great way for startups to transact with enterprise customers—we have a portfolio development team that is solely focused on opportunities like this, helping our portfolio companies join the marketplace and get ready to sell to enterprises, get through procurement and generally navigate the opportunities to partner with Microsoft and our customers.

That collaboration is what can turn breakthrough technologies into lasting impact, and it’s why partnerships are so critical in this next chapter of AI. 

Ready to learn more? Discover resources and tools to accelerate your AI journey

• Learn more about M12, Microsoft’s venture fund
• Explore cloud solutions, AI apps, and agents on Microsoft Marketplace
• Get the Prompt Guide for Business Leaders to turn AI intent into clear, actionable decisions
• Discover how Microsoft AI brings intelligence into the flow of work

The post FYAI: Why startups will help accelerate global AI transformation appeared first on The Microsoft Cloud Blog.

Frontier leaders aren’t simply bolting new technology onto their existing operations. As Microsoft Chief Executive Officer of Commercial Business Judson Althoff has shared in recent articles and keynotes, these leaders are taking a human-centered approach to AI transformation. The people closest to the work understand the real bottlenecks and opportunities. By equipping them with AI, leaders unlock practical solutions that drive measurable performance gains.

3 essentials for building a Frontier organization

Here’s Althoff’s outline for a Frontier approach to using AI and agents that puts capability directly into each employee’s hands.

1. Start with your employees to amplify ambition

At the heart of every Frontier business flow is the notion of democratizing intelligence. Human ambition is at the core, coupled with your AI assistants and agents to get real work done.

—Judson Althoff, Chief Executive Officer of Commercial Business, Microsoft

The idea: The point isn’t to simply deploy more technology, but to deploy it in ways that unlock more potential in people to solve their hardest problems and create more business impact. 

Why it matters: According to IDC,¹ AI adoption is accelerating past the initial experimentation phase, with 68% of companies using AI and 37% using agents. However, providing access to the technology is not the same as providing the guidance and skilling needed to unlock its potential. 

The shift: Frontier leaders focus on applying agents where they matter most—the priority workflows that define performance and growth. AI is at its most powerful when employees have the space and the guidance they need to imagine, experiment, and pursue bolder ideas. 

The big picture: Frontier leaders don’t start with AI capabilities. They start with human ambition, then design the systems, workflows, and guardrails that allow that ambition to scale responsibly. This requires treating AI adoption as a management system—not an IT rollout—with executives and business decision-makers actively redesigning workflows end-to-end.

2. Expand across every business function

There’s a maker in every one of us, and the Frontier Firm has a maker in every room of the house.

—Judson Althoff, Chief Executive Officer of Commercial Business, Microsoft

The idea: The people closest to the challenge are often closest to the opportunity. As AI becomes more accessible, creativity moves from the edges of the organization to the center so that everyone is empowered to innovate.

A striking data point: Frontier Firms aren’t leaving AI adoption to the IT department—they are making it a company-wide leadership priority. According to IDC research, Frontier Firms are using the technology across seven business functions on average.

Real-world innovations: Mercedes-Benz scaled AI innovation across its global production network, diagnosing efficiency declines and reducing energy consumption of buildings and machines—including 20% energy savings in one paint shop. And Althoff highlights how Toyota is pioneering AI intelligence in manufacturing with the O-beya system, a multi-agent AI system that simulates expert discussions virtually. O-beya can auto-select AI agents in fields like fuel efficiency, along with drivability, noise and vibration, energy management, and power management to pinpoint causes and suggest solutions. 

The takeaway: Broadening access to agents can unleash innovation. Frontier leaders don’t need to script how employees should use the technology—they just need to ensure that there are proper guardrails around a wide space for experimentation.

3. Trust, governance, and integration determine ROI

The idea: AI can create more value when people trust it enough to use it. Trust is what allows AI-powered innovation to scale beyond isolated pilots. And that requires human oversight with “observability at every layer of the stack,” according to Althoff. 

The challenge: Not every organization has put the right safeguards in place yet. Microsoft’s 2026 Data Security Index reports that only 47% of companies have fully implemented data security controls for AI.   

The solution: Frontier leaders must ensure security and be explicit about human-in-the-loop observability as a cornerstone of transformation. People adopt AI confidently when they understand how decisions are made, how data flows, and how systems behave—and when to intervene as needed. Finally, Frontier organizations don’t implement new technology and then slow down—or backtrack—to implement responsible practices. They design for trust from the start so they can keep moving quickly. 

Actions you can take to drive measurable impact

The idea: The organizations that will win in the Frontier era are those that view AI not as a one-off tool rollout but as a leadership discipline. They start by clarifying ambition, giving people the space and agency to act, and building trust early so transformation can scale across the business. Importantly, they use AI themselves to guide decisions, surface insights, and stress test ideas about keeping humans at the center of their business transformation. 

Where to start: Microsoft’s new Prompt Guide for Business Leaders was designed to help leaders get a handle on the changing AI landscape and use the technology itself to stress test their ideas and strategies in response to it. The guide offers guidance on how to:  

1. Assess readiness
2. Identify value 
3. Map workflows 
4. Build roadmap 
5. Plan for risk 
6. Define actionable next steps

Example prompt: “Show me the top three workflows where agents could reduce cycle time by at least 20% based on our current operations.”

From vision to value in the Frontier era

The guide demonstrates how AI can be a thinking partner, and helps leaders develop a strategy to help their people harness the technology to achieve goals, innovate, and unlock more value.

Innovation with AI

What every company can learn from Frontier Firms leading the AI revolution

Read the report

¹ IDC InfoBrief: sponsored by Microsoft, What Every Company Can Learn From Frontier Firms Leading the AI Revolution, IDC # US53838325, November 2025.

The post How to start your Frontier Transformation: 3 strategies to start with people appeared first on The Microsoft Cloud Blog.

Today, a wide range of advanced carbon dioxide removal projects are underway – and accelerating.

In Europe, construction is underway to enable carbon to be captured and condensed by Stockholm Exergi so it can be stored, beginning in 2028, deep under the North Sea. In the US, Lithos Carbon is spreading finely crushed volcanic rock across farmlands to convert carbon into bicarbonate ions that are stored for millennia deep in the ocean. And in Bolivia, Exomad Green is locking up carbon by processing sustainable forestry residues into biochar, a soil enhancement, using a type of thermal decomposition that was developed to break down organic materials in the absence of oxygen.

The goal is to help restore carbon balance lost through carbon dioxide emissions generated by modern life – from agriculture and construction to chemical and energy production.

Elementary school children are taught that plants remove carbon dioxide from the air through photosynthesis, so it’s no surprise that initial carbon removal management efforts, beginning in the 1970s, focused mainly on planting and conserving trees. Today, scientists have much more insight into what generates carbon dioxide emissions and alternative methods to remove them.

Building on that knowledge, a growing number of carbon dioxide removal firms around the world are developing innovative projects, and Microsoft is helping build a market for those solutions at larger scale.

In fiscal year 2025, Microsoft signed agreements to remove a record 45 million metric tonnes of carbon dioxide with 21 companies across the globe. That’s the equivalent of removing 9.8 million internal combustion cars from the road for a year.

Since the launch of Microsoft’s carbon dioxide removal program, the pace and scale of projects have continually accelerated. Microsoft’s carbon dioxide removal agreements in fiscal year 2025 are two times the volume contracted in the preceding fiscal year and nine times the volume contracted in fiscal year 2023. Projects span the globe, including ones in the US, Brazil, Denmark, Sweden, Bolivia, Norway, India, Panama, Canada and Switzerland.

And while this sharp increase in advance purchase commitments indicates rapid market advancement, there is a long road ahead to get to the seven to nine billion tonnes of carbon dioxide removal that experts say the world needs annually by 2050.

Microsoft has committed to become carbon negative by 2030 and to remove its historic emissions by 2050. The company takes a reduction-first approach. Carbon reduction is the most powerful lever for achieving carbon negative because it directly lowers emissions at the source, said Phil Goodman, director of the carbon removal portfolio at Microsoft. From there, carbon removal provides a direct pathway to address residual and historical emissions that cannot be mitigated through reduction. Microsoft advances carbon dioxide removal by signing contracts with carbon removal project developers to deliver volumes of high-quality carbon dioxide removal.

Key to this process is the carbon removal credit, representing the removal of one metric tonne of carbon dioxide. Credits are listed by project developers on registries, where they are bought by companies seeking to advance carbon markets and meet net-zero goals. Registration requires independent auditors to verify carbon removal claims according to the registry’s methodology. Registries retain a public ledger to ensure transparency and prevent double counting so the credit cannot be used twice.

However, due to the small size of the carbon market, in the past several years Microsoft increasingly has worked directly with carbon dioxide removal suppliers during early stages of project design and development before they list their credits on registries. After conducting extensive due diligence, Microsoft invests to help develop innovation by signing contracts with carbon removal project developers.

“With any form of carbon removal, you need someone out there to buy the credits so that the economic model works,” said Goodman. “By securing that forward demand commitment, suppliers can actually go raise financing, hire staff and build out their projects. We buy only a fraction of a project’s total credits, and we hope other companies can make faster procurement decisions knowing that projects in our portfolio underwent deep due diligence.”

Measurement is the essential backbone

Growing awareness of the harms of excess carbon dioxide in the atmosphere has spurred scientists to develop increasingly sophisticated methods for removing it and storing it safely – in soil, geologic formations, in trees and in the ocean.

Despite progress, carbon markets have also faced setbacks. In the last several years, scrutiny of credits generated by forest preservation projects helped lead to a regulatory crackdown.

To ensure the integrity of its program, Microsoft applies a rigorous due diligence process. Potential suppliers must submit comprehensive documentation detailing strategies for project design and implementation, methodologies, validation processes and other essential plans for addressing Microsoft’s quality criteria.

“Measurement is the essential backbone of this industry, because there needs to be a common understanding of what a carbon credit means,” said Goodman. “You’re not buying a physical good. You’re buying this environmental attribute that you need to feel confident can be consistently applied across project types, apples to apples.”

Microsoft takes what it calls a “portfolio” approach, spanning a combination of nature-based and engineered solutions, which it believes is needed to catalyze market development, said Goodman. The company evaluates projects based on durability – or the length of time that carbon dioxide is expected to remain out of the atmosphere – including some that are expected to store carbon dioxide for less than 100 years to 1,000 years or longer.

Projects must show they will remove more carbon dioxide than they emit, to ensure that the net removal accounts for all operational and embodied emissions, and that local communities and ecosystems won’t be negatively impacted by other parts of the process.

Enhanced rock weathering: Captured on the farm for storage in the ocean

Measurement is at the heart of Lithos Carbon, co-founded by two academics from Yale and Georgia Tech who helped pioneer monitoring, risk mitigation and verification of a process known as enhanced rock weathering to remove carbon dioxide and improve soil health.

The process begins with the sourcing of finely ground basalt, a type of volcanic rock that is widely available as a waste byproduct of quarrying and already used in composite materials for construction, roads and in other industries. Carbon dioxide removal depends on the chemical reactions that occur after the ground rock is spread over farmland, where it mixes with rainwater and atmospheric carbon to create stable bicarbonate ions. These ions then seep into waterways and rivers before heading out to sea, where they are stored for 10,000 years or more.

Lithos’ approach accelerates and condenses a process that normally takes tens of thousands of years into a few seasons. “Essentially, we’re helping nature do something it already does – just much faster,” said Chief Executive Officer Mary Yap. Following an initial pilot project, Lithos and Microsoft signed a three-year contract in 2024 to remove over 11,400 metric tonnes of carbon dioxide.

Lithos uses high-density soil sampling and advanced laboratory analysis to make the invisible elements in soil visible, followed by AI-powered predictive modeling tools, water runoff samples and natural water chemistry to determine the end-to-end flow of bicarbonate ions from farms to waterways. Microsoft also assisted Lithos in creating a new, highly sensitive technique for detecting bicarbonate in rivers that originates from crushed basalt. Collaborating with Microsoft “helped accelerate the science and operational discipline that will help define the future of our work in this field,” said Yap.

For farmers, the benefit is improved soil health, which leads to higher crop yields. “We roll up with a bunch of trucks,” said Yap. “We spread this volcanic, nutrient-rich dust that restores their soils, reduces soil erosion and replenishes essential nutrients. It helps both farmers and the planet at the same time.”

Regenerative farming: Capturing carbon while improving soil and crop yields

Changing farming and ranching practices offer other avenues for addressing and removing global greenhouse gas emissions. Excessive fertilizer use can lead to nitrous oxide emissions, a powerful greenhouse gas, while traditional tilling accelerates the oxidation of soil organic carbon, releasing it into the atmosphere as carbon dioxide. Tractors and other farm machinery burn fossil fuel. Given its size and impact, agriculture “is a sector that just cannot be ignored,” said Dylan Lubbe, commercial director at Agoro Carbon Alliance, which works with the agricultural sector to store carbon.

To attack the problem, Agoro has recruited a small army of agronomists and other agriculture experts to help an expanding network of farmers and ranchers move away from traditional agricultural practices in favor of regenerative farming and ranching, a practice that removes carbon, improves soil health and boosts crop yields. It includes no-till soil preparation and the use of clover and other so-called cover crops with deep root systems that transfer atmospheric carbon into the soil through root biomass and microbial processes. Ranchers are encouraged to rotate animals to avoid over-grazing and stimulate plant growth.  

“Soil is a natural carbon sink,” said Lubbe. “Traditional agricultural practices disturb the soil, essentially letting that stored carbon dissipate into the atmosphere.”

In June 2025, Microsoft reached an agreement with Agoro to purchase 2.6 million metric tonnes of carbon dioxide removals over 12 years. Extensive soil sampling validated by third parties is expected to confirm the project’s carbon removal and storage and soil enhancement.

Agoro provides financial support for farmers and ranchers so they can feel more comfortable that changing to new farming and ranching methods won’t hurt their bottom line. “It’s very important to have a for-profit or market mechanism approach to a problem,” said Lubbe. “If there’s money to be made, then people are incentivized to address something.”

BECCS: Bioenergy with carbon capture and geological storage  

Another way to remove carbon dioxide is to capture it from burning biomass and store it underground in geological formations, using know-how first developed by the oil and gas industry. Stockholm Exergi, a Swedish energy giant, is addressing the complexity and high upfront costs of this approach by retrofitting its enormous bioenergy plant on the Stockholm waterfront so carbon dioxide can be captured by burning forestry waste from sustainably managed harvests.

Besides generating electricity, the plant already uses heat generated by its bioenergy process to warm the city of Stockholm. Now construction is underway to make it possible for carbon dioxide to be separated from flue gases and condensed so it can be transported by ship to the North Sea for storage. It will then be injected more than 800 meters beneath the ocean floor into porous bedrock, where it mineralizes over time. This elaborate process is known as bioenergy with carbon capture and storage, or BECCS.

“We are creating a flow of carbon from the atmosphere to the geosphere,” said Erik Rylander, head of commercial operations at Stockholm Exergi.

In May 2025, Microsoft reached an expanded ten-year agreement with Stockholm Exergi to remove over 5 million metric tonnes of carbon dioxide. “This has never been done before at this scale and wouldn’t have been possible without their commitment to buying our product,” said Rylander. Working with Microsoft, “we are really trying to develop something that can become a market standard for BECCS.”

Tropical reforestation: Drones and satellites replace tape measures

Harnessing trees to capture carbon dioxide has come a long way since the 1980s. While measuring the width of tree trunks with tape measures still plays an important role in the field, as was typical of early reforestation efforts, Brazil’s re.green increasingly uses light-emitting drones and satellite imagery to monitor forest growth and carbon removal. Drones also distribute fertilizer and seeds over remote areas and steep hills. Satellite images and machine learning models help identify prime areas for reforestation.

To lock up carbon dioxide, re.green operates as an ecological restoration company, reclaiming lands that were deforested decades or even centuries ago. “Our effort is to restore the original ecosystem,” said chief executive Thiago Picolo. “We’re not just planting rows of trees.” Their work ranges from selecting the most suitable native species for each landscape to sowing a diverse mix of seedlings to removing invasive exotic grasses and controlling ants that attack seeds and seedlings.

In January 2025, re.green and Microsoft reached an expanded agreement to purchase a total of nearly 6.5 million metric tonnes of carbon dioxide and restore a total of 33,000 hectares in the Amazon and Atlantic Forest biomes. “Scale is key to biodiversity, resilience and efficiency,” said Picolo.

While trees consume carbon dioxide as they grow, they also release it as they decay, die or catch fire. Because their carbon-absorbing capacity also plateaus over time, re.green strategically harvests select native timber to maintain forest health for carbon capture and enhance biodiversity. While planting and sustainable harvesting also create local jobs, the priority is supporting the restoration of a healthy tropical forest.

“My dream, my Holy Grail, is not only for re.green to be a successful company, but for us to be a leader in a thriving, sustainable, new economic sector,” said Picolo.

Top image: While planting and sustainable harvesting creates local jobs, re.green’s priority is restoring healthy tropical forests. Photo provided by re.green.

Related links:

Learn more: Microsoft’s carbon dioxide removal portfolio

Learn more: Microsoft’s carbon dioxide removal program

Learn more: Microsoft’s sustainability commitments

Sally Beatty writes about AI and innovation at Microsoft, focusing on the company’s most cutting-edge breakthroughs and how emerging technologies are improving the lives of ordinary people. Previously, Sally was a feature reporter for the Wall Street Journal in New York, where she broke news about media and marketing. A California native, Sally lived in Hong Kong in the early 1990s, where she wrote for the Journal about the societal impact of economic and political reforms and consumer spending trends. You can contact Sally at LinkedIn.

The post From farms to oceans: How Microsoft is working to scale carbon dioxide removal appeared first on The Microsoft Cloud Blog.

Today’s business environment demands more with less. Organizations must deliver better personalization, higher volume, and increasingly complex insights while operating with greater efficiency. The gap between stakeholder expectations and what teams can realistically deliver continues to widen. 

Microsoft’s recent insights on Frontier Transformation address these challenges by embedding AI into the core of operations. Frontier Firms are organizations that treat AI as a foundational capability and are already transforming how they work. 

Frontier Firms don’t simply automate; they adapt. By adding adaptive intelligence to existing systems, they unlock three advantages: 

• Awareness: Systems perceive conditions in real time. 
• Reasoning: They prioritize tasks based on business needs. 
• Interaction: They communicate in natural, intuitive ways. 

Early adopters see small improvements compound quickly. These include faster service, more accurate recommendations, fewer equipment surprises, and clearer insights into peak times and bottlenecks. As agentic AI matures, companies can offer guidance and assistance that feels intuitive. Employees gain more time for high-value work, and leaders gain deeper visibility into operations. 

Frontier Transformation is more than a technology upgrade. It represents a shift in operating model. Organizations that treat AI as a foundation will lead the next wave of business innovation. 

Agentic AI is reshaping customer experience 

This shift is already visible in retail, where agentic robots are transforming customer experience and improving operational performance. Customers expect fast, personalized service, yet retailers often face staffing constraints, training gaps, and unpredictable demand. 

Industry studies show: 

• 75% of consumers are more likely to purchase when recommendations feel relevant. 
• Nearly 40% of in-store complaints relate to wait times. 
• Inventory inaccuracies account for 4–8% of lost sales. 

These challenges reflect a broader pattern across frontline-heavy industries. Customer expectations continue to rise, and employee workloads grow more complex. 

Microsoft’s Work Trend Index reinforces this dynamic. Frontline employees say AI tools that reduce repetitive tasks, surface real-time information, and streamline customer interactions have the biggest impact on satisfaction and performance. As organizations integrate adaptive intelligence into daily workflows, these benefits build on each other and help accelerate Frontier Transformation. 

Recent industry research shows that retail and consumer packaged goods organizations are generating significant business value from generative and agentic AI, with early deployments consistently delivering multi-times ROI and accelerating impact across frontline operations.

Agentic AI creates new possibilities for stores. Instead of relying on rigid automation, it blends environmental awareness, adaptive reasoning, and conversational interaction to help teams respond in real time. 

ADAM: From beverage service to customer care 

Richtech Robotics’ ADAM beverage robot illustrates how quickly agentic systems can enhance the customer experience. Richtech, based in Las Vegas, designs and commercializes autonomous robotic solutions for hospitality, retail, logistics, and manufacturing. Through a close, hands-on collaboration between Richtech’s engineering team and the Microsoft AI Co-Innovation Labs, the two companies jointly developed new adaptive intelligence for ADAM—transforming it into a conversational, context-aware assistant powered by Microsoft Azure AI. These enhancements enabled ADAM to move beyond routine beverage preparation and support richer customer interactions.

Today, ADAM: 

• Adjusts recommendations based on weather, time of day, and promotions. 
• Responds naturally to customer requests like “less sweet,” “extra ice,” or “what’s seasonal?” 
• Notifies staff about ingredient or equipment issues before problems occur. 
• Uses vision models to maintain speed and quality during busy periods. 

Retailers report smoother operations and better customer feedback. ADAM is context aware, conversational, and reliable—qualities customers consistently reward and areas where AI has historically struggled. 

While ADAM is a retail example, the pattern extends far beyond beverage automation. Across logistics, healthcare, hospitality, and manufacturing, Frontier Firms are adding ambient intelligence and agentic workflows to physical operations and seeing meaningful gains as a result. 

Unlocking retail transformation at scale 

Once retailers see how intelligence enhances a single customer interaction, the next question naturally follows: where else can this help? Building on the advancements made with ADAM, Richtech Robotics is extending these capabilities through its Agentic Store initiative. By applying vision, voice, and agentic reasoning to common in-store tasks, the initiative helps retailers address friction points that slow down the shopping experience. 

Examples under development include: 

• Robots that guide customers to products.
• Systems that detect empty shelves or misplaced items.
• Voice-enabled, in-aisle assistance.
• Real-time adjustments based on foot traffic or local events.

This approach does not require heavy hardware investments. These workflows are software-driven and build on existing store infrastructure. It reflects how Frontier Firms drive transformation by spreading intelligence across the ecosystem rather than upgrading a single process at a time. 

Retailers gain clearer visibility into peak demand, customer behavior, product movement, and service quality without increasing manual tracking. As one store manager described it, “it feels like having a second set of eyes that never gets tired.” 

Convenient, high-quality service becomes a blueprint for store-wide intelligence. In the coming years, a clear difference will emerge between retailers that treat AI as a tool and those that treat it as a foundation. The latter will set the pace for the industry. 

Steps toward Frontier Transformation 

Agentic AI gives retailers a practical and achievable path forward. It elevates customer experience, reduces operational strain, and creates the foundation for smarter and more adaptive stores. Organizations that embrace Frontier Transformation position themselves as Frontier Firms, ready to scale faster, work more intelligently, and unlock new value through the combination of human judgment and AI-driven insight. 

The journey begins with small, strategic steps and a bold vision for what is possible. To explore the broader business impact of AI across frontline and customer-facing roles, review Microsoft’s Work Trend Index: The year the Frontier Firm is born.

Explore how organizations are transforming with AI, and learn how you can build your own generative AI proof of concept with the Microsoft AI Co-Innovation Labs.

The post Frontier Transformation in retail: How agentic AI robots are redefining store experiences appeared first on The Microsoft Cloud Blog.