The more time I spend working with AI, the more convinced I become that what matters most for her future isn’t how quickly she can answer. It’s whether she has the judgment to know when an answer can be trusted.

I’ve spent decades at Microsoft watching this tension play out: first building tools for other developers, then working across AI as models moved from research curiosities to systems deployed at scale. Now we’re building Microsoft IQ, where we’re exploring how an organization’s collective intelligence can become its greatest advantage. Through every one of those chapters, one thing has remained true: it’s never enough for a system to be powerful; it must also be trustworthy.

Trust is what turns assistance into delegation. When we can trust an agent to do what we intend, within the limits we set, we can hand off the work we never wanted to spend our lives on: the repetitive tasks that drain attention, the mundane work that fills a day without moving anything meaningful forward, the dangerous work humans should not have to do, the work too vast for any individual or team. Agents should take on that toil, extend our reach, and give us back our time for the work that calls for something only humans bring.

My daughter doesn’t know any of this yet. But by the time she’s grown, most of the work that rewards speed and repetition will be work we delegate. What will matter then is exactly what gave her teachers pause: the patience to stay with a hard problem, reason through it, and decide when she’s reached a conclusion she can trust. The very thing they feared might hold her back could be exactly what the next era prizes most.

So no, I’m not worried about the timed test. I hope she grows up in a world where software carries the toil and people are freed for the work that is unmistakably ours—to think, to judge, to create, to care for one another. That is the future I want agents to make real. But my hope is not evidence it will happen. The future I just described depends on a single question: can we trust agents to do the work? Trust is earned one task at a time. So, I went looking for evidence of where it’s been earned, and where it hasn’t.

We partnered with MIT Technology Review Insights on new research that draws directly from the technical leaders building this frontier: not the people talking about it, but the people doing it. We surveyed 300 technical experts across AI, data, and cloud domains, spanning 12 industries and 4 regions of the world, asking them to rank their confidence across 101 of the top tasks. What we got back is the 2026 Agent Confidence Index, an honest map of where agents are delivering real value, so our community can see what’s working and move forward together with conviction.

Learn from where confidence is highest

Across the 101 tasks measured, average confidence already lands at 64 out of 100, and thirty tasks clear 70. The highest scores cluster on work that is both predictable and draining: the late nights, the interruptions, the low-value repetition. Automated report generation leads at 83.5. Boilerplate code generation for new features sits at 82.5—the hours a developer no longer spends rewriting the same patterns, freed for the work that challenges them. Certificate expiration monitoring and renewal, at 81.5, ends the scramble that pulls engineers off high-stakes problems for something entirely routine. Real-time data stream monitoring follows at 80.5, and release note generation from commit history at 79.5—the manual end-of-sprint commit review, gone. This is where frontier teams are already delegating to agents, regularly.

The pattern holds across every discipline. In developer and AI workflows it extends to API client maintenance and code identification; in cloud operations, to ticket routing and cost optimization; in data, to anomaly detection. Wherever it sits in the stack, this is work technical teams now trust agents to own.

What matters most here isn’t what the data says about the tasks; it’s what it says about the people delegating them. When technical experts believe in something deeply enough to hand it real work, that belief ripples outward. It becomes the recommendation they make to their leadership, the solution they build for their customers, and the culture they create for their teams.

Even the toughest agent tasks are gaining traction

Here’s what strikes me most: the tasks ranked lower on the index are still high in absolute terms. Service mesh configuration and troubleshooting sits at 37.5, database schema migration scripting at 46.5, memory leak detection at 48.5. These sit at the very frontier, the interconnected, high-stakes work where investment and innovation are concentrated right now.

Consider what they demand. Service mesh configuration touches many systems at once. Database migration carries real stakes, requiring precision across data, application, and infrastructure layers at the same time. Memory leak detection means diving deep into a system’s behavior under load, accounting for conditions that shift from one deployment to the next. These are the challenges that have separated great engineers from exceptional ones—and even here, experts see agents helping. Not carrying the work alone, but contributing where it used to be unthinkable. That confidence is still climbing, and that’s telling.

We’re shipping new capabilities constantly to support this momentum. Database migration tooling in GitHub Copilot now covers not just scripts but the full application and infrastructure migration story. The Azure Site Reliability Engineering (SRE) Agent brings decades of experience operating Azure at scale and deep profiling capabilities directly into memory analysis and performance diagnosis.

Why human judgment remains paramount

When we asked technical experts how they’re navigating agent adoption, 59% named “keeping humans in the loop” as their top priority—ahead of better observability, ahead of governance documentation, and ahead of everything else. That’s a mark of maturity. Teams moving forward with clarity treat agent oversight as non-negotiable, regardless of how capabilities evolve.

The boundary itself is straightforward. Agents excel at well-specified, high-volume, reversible work: they synthesize data, automate known workflows, and surface anomalies at a speed and scale no human team could match. The moment a decision becomes high-stakes, context-dependent, or hard to undo, a human signs off. That isn’t a limitation of the technology; it’s the architecture of a trustworthy system.

What’s changing, and what remains underappreciated, is the skill it takes to draw that boundary well: the discipline of full-lifecycle evaluations and guardrails. Success means measuring agent output against intent and keeping behavior inside your business strategy. It’s new territory for most engineering teams, and it’s becoming table stakes for modern software faster than most organizations realize. The good news: the same tools generating the work can help you build the harness. Ask GitHub Copilot to write the evals and it will. Frontier teams are already doing this, and it’s why they’re pulling ahead.

Agents are opening career doors for engineering

Across system reliability and site operations, evaluations and quality assurance, and data pipeline management, 80% or more of respondents see meaningful career opportunity ahead. We believe this is one of the most significant moments in the history of building software, not because agents replace what technical people do, but because what’s left when they take on the toil is the work that defines a career: the judgment calls, the architectural vision, the reasoning to navigate complexity under pressure. That fluency will define the next generation of technical leadership.

We’re living this shift at Microsoft, right alongside our customers. Junior developers are using agents to explore codebases on their own and arriving at mentoring conversations with sharper, more sophisticated questions. Senior engineers are covering more ground because the repetitive work that used to fill their days is now delegated, and the work that’s left is harder, interesting, and consequential. Both are growing into more capable versions of themselves. For me, that’s the outcome I’ve always believed technology could deliver.

An integrated approach to intelligence and trust

Designing more sophisticated agent systems has made one thing clear: agents thrive in well-integrated environments, working best when your whole stack draws on a single source of truth. The high-confidence tasks are the ones we’ve already figured out; the meaningful frontier is the harder, interconnected work, and that’s exactly where observability, governance, security, and unified intelligence have to operate as one.

Microsoft IQ brings your enterprise context into a single, continuous intelligence layer. Within it, Work IQ builds semantic understanding of how your business operates across email, calendar, meetings, chats, files, people, and collaboration patterns. Such depth of knowledge is the reason technical teams choose us, and it’s what drives my focus and passion in learning how people actually work so their agents get them. My colleague Kim Manis, CVP of Product for Microsoft Fabric, has written specifically about what this means for data professionals, and the integral role of Fabric IQ.

It’s all part of the Microsoft Agent Platform, which is becoming the operating system for enterprise AI at scale. From building in GitHub and contextualizing with Microsoft IQ, to running in Microsoft Foundry and governing in Microsoft Agent 365, Microsoft is uniquely positioned to help customers bring together data, models, agents, and human judgment into a continuously improving and secure system.

Frontier transformation is being led by builders like you.

Next steps:

• Download The 2026 Agent Confidence Index from our partners at MIT Technology Review Insights. It is a free, ungated deep dive into all 101 tasks, broken out by role and workflow, with the patterns and reasoning behind where confidence is strongest and the frontier is expanding.

• Join us at the AI Engineering World’s Fair (June 29-July 2) where our very own Pablo Castro will keynote, and our teams will offer 16 breakout sessions and 4 labs. Swing by the Microsoft booth as well to explore an interactive 3D visualization of the Index data. We want to hear what’s working for you.

• Learn more about Microsoft IQ and how it connects across Work IQ, Fabric IQ, Foundry IQ, and the newly announced Web IQ. You can catch up on all the developer innovation from Microsoft Build through Satya Nadella’s keynote, Kyle Daigle’s blog post, and the Microsoft Build CLI.

What’s Working in Agentic AI

The 2026 Agent Confidence Index report reveals where agents are trusted, the challenges they face, and what leaders should do next

Download the 2026 Agent Confidence Index

The post The 2026 Agent Confidence Index: Where 300 builders see real momentum appeared first on The Microsoft Cloud Blog.

As operators move from experimentation to execution, the opportunity is not just to adopt more AI, but to build systems that combine intelligence with trust across networks, data, and customer operations. TM Forum DTW Ignite 2026 focuses on how those capabilities come together to drive measurable business outcomes.

Composable IT and ecosystems

Complexity continues to slow progress across telecom. Tightly coupled systems, fragmented integration paths, and years of technical debt make it harder to launch new services, adapt quickly, and deliver consistent experiences across channels.

TM Forum continues to position Open Digital Architecture as both a practical migration path away from legacy complexity and the architectural foundation for AI-native operations, with tools, standards, Open APIs, and reusable components designed to support more modular, cloud-native, and interoperable platforms.

That is why the Composable IT and Ecosystems summit sits so centrally in this year’s event: it reflects a broader industry shift from digital transformation as a long-running aspiration to reinvention for the AI era, where openness, composability, and governance become prerequisites for scaling intelligent operations.

For operators, the value is practical. Modular environments make it easier to modernize in stages, improve interoperability across business and network domains, and create a stronger foundation for ecosystem plays such as CAMARA and Open Gateway. This is where solutions such as Infobip’s Network API offering become relevant—helping operators turn standardized capabilities like number verification, SIM swap detection, and device location into practical services that expand revenues, strengthen trust, reduce fraud, and reach more developers.

With Microsoft, teams can standardize platforms, simplify integration, and create an environment where new capabilities can be introduced with greater speed and confidence.

That conversation will continue in the panel discussion “The AI-Native ODA Roadmap.” The session will explore how operators can build AI-native operations on common foundations across IT and networks, with emphasis on interoperability, governance, reusable patterns, and business value. It is a timely example of how TM Forum DTW Ignite 2026 is connecting architectural change to the strategic and operational priorities telecom leaders are navigating now.

That same modernization path is also reshaping the business side of telecom through Agentic Business Support Systems (BSS). Instead of hard-coded workflows, operators can use AI agents to interpret intent, orchestrate actions across catalog, ordering, billing, and care, and accelerate the path from request to resolution.

With Microsoft, TM Forum Open APIs, and a composable architecture, BSS can evolve into a more adaptive, outcome-driven system that enables faster service creation, stronger customer experiences, and new monetization opportunities across the ecosystem.

This direction is reinforced by recent Microsoft Build 2026 announcements, where capabilities such as Work IQ and Agent 365 introduce a consistent way to ground AI agents in enterprise context, apply governance by design, and orchestrate actions across systems. Together, these capabilities make it easier for operators to move from isolated automation to agent-driven workflows that are composable, secure, and production-ready at scale.

This evolution also reflects the emergence of more persistent, context-aware agents—exemplified by experiences such as Microsoft Scout—that can operate continuously across tasks and systems, moving from reactive copilots to proactive digital coworkers that help coordinate work across the enterprise.

Autonomous networks

The industry is moving beyond manual processes toward systems that can interpret signals, recommend next actions, and automate response within clear controls. That shift sits at the heart of the Autonomous Networks summit at DTW Ignite 2026, where the focus is on how operators can make autonomy practical in day-to-day operations.

AT&T offers a strong example for incident management. By improving observability, creating better context across operational data, and accelerating the path from detection to action, operators can improve reliability and shorten the time between issue discovery and resolution. Microsoft supports this direction with cloud-scale data services, AI capabilities, and a platform approach that helps teams operationalize intelligence rather than isolate it in pilots.

 Microsoft’s Network Operations Agent (NOA) framework extends this vision by enabling operators to deploy coordinated AI agents across network operations—helping interpret signals, orchestrate actions, and drive closed-loop automation with greater consistency and control.

The broader direction is clear. Operators are moving toward more predictive operations, more intelligent assurance, and more trusted automation that improves resilience, experience, and efficiency.

AT&T will also bring this story to the Park stage at DTW Ignite 2026, where the discussion is expected to highlight how End-to-End Incident Management uses current context, observability, and AI to move faster from detection to action. It adds a timely, customer-led example of how operators are applying agentic decision-making to improve reliability, reduce disruption, and make autonomy more practical in live network operations.

By bringing more context and intelligence into incident management, we are improving how quickly teams can identify issues, coordinate response, and restore service with confidence. This type of advanced agentic technology is helping us keep our customers connected and provide a better experience for them.

—Mark Austin, Vice President Data Science, AT&T

Underpinning this shift is the emergence of unified data and AI platforms that can ground agents in real-time operational context. With Microsoft Foundry and Foundry IQ, its knowledge layer, operators can connect data across network, service, and customer domains, enabling agents to reason more effectively and support closed-loop decisioning that moves from insight to action with greater speed and consistency.

Microsoft’s recent work in Azure Networking operations adds an important operational lens to this shift. Rather than treating agents as simple assistants, the model is evolving toward more persistent digital coworkers that help coordinate the messy middle of operations across incidents, repairs, vendors, and validation steps. In large-scale environments, that can reduce manual effort, shorten mitigation time, and free engineers to focus more on judgment, oversight, and higher-value decisions while agents keep execution moving across systems and teams.

Trustworthy AI and data

AI creates value only when it is grounded in trusted data, strong governance, and a clear path to action. In telecom, that standard matters because operators work in highly regulated environments and depend on reliable decision-making across critical services.

That is why the Trustworthy AI and Data summit at DTW Ignite 2026 matters. The conversation is not only about model capability. It is about data readiness, governance, observability, and the controls required to scale AI responsibly across the business.

This is reinforced by Microsoft’s perspective on local AI, where operators extend the AI execution model closer to where data is generated—improving latency, resiliency, and control while keeping data private. Running inference locally or at the edge can also shift the cost profile, reducing dependency on centralized compute for time-sensitive or high-volume workloads while enabling new real-time scenarios.

SoftBank brings that point into focus through its AI-powered customer platform, where the value shows up in responsiveness, service quality, and workload reduction. In its work with Microsoft, SoftBank is applying Microsoft Foundry, Azure OpenAI in Foundry Models, Azure AI Search, and Semantic Kernel to help build a more intelligent call center experience that reduces wait times, improves response quality, and supports more scalable customer operations. The example reinforces that AI success depends on the right data foundation and a disciplined link to measurable outcomes.

Microsoft sees telecom ontologies as an important part of the foundation for trustworthy AI. By creating a shared semantic model across network, service, customer, and operations data, they help AI systems reason with more context, improve explainability, and support more consistent decisions across domains. Combined with Microsoft Fabric as the unified data foundation, Fabric IQ as the intelligence layer, and Microsoft’s broader data and AI platform, that structure can help operators reduce fragmentation and move with greater confidence from isolated use cases to governed, reusable intelligence.

That theme will also come through in Alberto Manuel Fernandes Dias’ panel, “Data products at scale—the blueprint for AI-native telcos.” The discussion reinforces that scaling AI in telecom depends on more than models alone. It requires well-designed data products, clear ownership, and disciplined governance to turn fragmented information into reusable, high-quality, and AI-ready assets. For operators, that blueprint matters because it creates the consistency, trust, and lifecycle discipline needed to support autonomy at speed and scale.

Accelerating transformation through partner innovation

Transformation in telecom does not happen in isolation. Across TM Forum DTW Ignite 2026, Microsoft and its partners are working together to help operators modernize networks, simplify operations, and bring AI into real business processes.

From composable IT and cloud-native BSS modernization to agentic marketplaces, digital twins, and voice AI pilots, these collaborations reflect how the industry is moving from experimentation to scalable, outcome-driven innovation.

• Amdocs is highlighting continued momentum in cloud-first modernization through its expanded work with Lumen, helping accelerate enterprise billing transformation on Microsoft Azure. Separately, Amdocs is also showcasing how its Entitlement Server on Azure is setting new performance benchmarks for digital service delivery, reinforcing the value of scalable, AI-ready cloud infrastructure for telecom providers.
• Beyond Now is also demonstrating how AI is reshaping telecom monetization through its work on agentic marketplace experiences. In collaboration with Microsoft, the company is highlighting how operators can use intelligent automation and AI-powered ecosystems to accelerate service innovation, streamline partner engagement, and unlock new digital revenue opportunities.
• Kenmei is advancing how operators can make network data more usable for AI-powered operations through its new Network Performance Data Product built on Microsoft Fabric. Running natively in Azure, the solution combines governed network KPIs with a telecom ontology layer, helping operators accelerate agent-ready use cases while keeping data secure, unified, and under their own control.
• By integrating Nokia Data Suite with Microsoft Fabric, operators can access on-demand telecom data products that simplify data integration across complex, multi-vendor environments. This unified data foundation enables faster adoption of AI, generative AI, and automation, supporting the journey toward autonomous networks.
• Norwood is demonstrating how voice AI is becoming more practical for telecom providers through its OpenSpan Voice platform. In collaboration with Microsoft, the company is piloting AI-powered small and mid-sized businesses (SMB) voice capabilities with a major United Kingdom and Australian telecommunications provider, demonstrating how operators can deliver more intelligent, responsive customer communication experiences at scale.
• Tech Mahindra and Microsoft are advancing telecom innovation with a 5G Network Digital Twin, an intelligent virtual replica that enables AI-powered optimization, proactive issue resolution, and new monetization opportunities. Built on Microsoft Azure and Fabric, the solution unifies data, simulates network scenarios, and drives measurable impact across cost efficiency, network slicing, service assurance, and autonomous operations, while maintaining enterprise-grade governance and data sovereignty.

Customer momentum: From blueprint to production

The strongest signal in telecom today is that customer stories are moving from blueprint to production. Leaders want to see where transformation is working in practice and where AI is delivering value beyond the pilot stage.

• MEO stands out because its modular AI factory connects strategy to execution, giving the company a more repeatable path to scale new use cases across the business. AT&T highlights the operational side of the story, with incident management illustrating how better data and more intelligent workflows can improve detection, triage, and resolution.
• TIM Brazil adds a security-focused example, showing how integrated signals and intelligent protection can strengthen operational resilience while reducing noise and improving response. Together, these examples show how operators are turning transformation into something more tangible, operational, and repeatable.

TM Forum also shows how quickly AI can move from experimentation to scaled impact. Using Microsoft 365 Copilot and Microsoft Copilot Studio, the organization moved rapidly from early pilots to production with internal and member-facing agents. Its cross-functional productivity agent, Buddy, is already delivering measurable ROI across multiple departments, while Navigator—launched in time for TM Forum DTW Ignite 2026—simplifies access to standards and articles for members. The result is a strong example of how AI can improve internal productivity, strengthen knowledge access, and create a foundation for broader support across the innovation cycle, from catalyst to production.

That momentum is visible across the telecom landscape. Recent public examples show operators using Copilot to strengthen leadership decision-making, workforce productivity, and sales execution—from Singtel and Ooredoo Qatar to KPN—while large-scale deployments across Accenture, Cognizant, Infosys, TCS, and Wipro reinforce how quickly the supporting ecosystem is scaling the skills, services, and implementation capacity needed to help enterprises and operators move from pilots to broader adoption. Together, they point to the same shift: AI is becoming part of day-to-day work, with measurable impact on speed, decision quality, and business outcomes.

That is why TM Forum DTW Ignite 2026 feels timely. Innovation matters, but adoption matters more. The next chapter of telecom will be shaped by operators that can turn transformation into repeatable outcomes.

Join us in Copenhagen

TM Forum DTW Ignite 2026 arrives at a pivotal moment for the telecom industry. TM Forum has structured the event around three mission summits all designed to help operators move from ambition to real change:

• Composable IT and Ecosystems
• Autonomous Networks
• Trustworthy AI and Data

We look forward to continuing these conversations in Copenhagen and sharing how Microsoft, together with customers and partners, is helping telecom organizations translate strategy into measurable progress. In the end, the next chapter of telecom will be shaped by operators that can pair intelligence with trust at scale.

Microsoft for telecommunications

Accelerate your business performance and growth with Microsoft AI

Discover more

The post TM Forum DTW Ignite 2026: From strategy to execution in the AI era of telecom appeared first on The Microsoft Cloud Blog.

1. Will AI amplify the intelligence of my organization and the attributes that make my company unique within its industry to grow my business; or will it use my intelligence for its own benefit, learning from my most important business flows and leveraging my intellectual property?
2. Can I trust that the outcomes are providing durable return on investment and that these solutions are running within the confines of my governance and security standards?
3. How do I get the visibility, control, flexibility and business model innovation needed to manage the costs associated with AI and maximize value?

I consistently advise customers that they need to build their own IQ on a platform of intelligence that is model-diverse, open and heterogeneous at every layer of the stack. Models are commoditizing. No company should be dependent upon any one model or any one model’s harness. Over the weekend, Satya warned of a world where every company across every sector is ceding value to a few models that eat everything they see. AI that is intended for growth should amplify the intelligence of an organization so that it compounds from within.

Companies also need an observability platform that provides governance, management, security and Financial Operations (FinOps) to ensure the ROI with AI. This enables AI to be trusted within the environment over which it reasons and puts the business in control of the outcomes.

Intelligence + Trust is embedded across Microsoft 365 Copilot, GitHub Copilot and Copilot Studio, where model diversity aligns cost and performance to each task. Microsoft IQ optimizes workflows, so context is routed efficiently and reduces unnecessary compute. Agent 365 is the control plane to observe, govern, manage and secure agents. We have built a system to manage AI spend as a core enterprise capability, not an afterthought. It is delivered across clouds and model providers without locking customers into a single approach.

Managing costs at scale

As agent usage scales, organizations need a clear set of levers to manage cost:

Model diversity. Any given inferencing model, model harness or agentic loop on its own does not help build out an organization’s IQ in ways that compound its intelligence. Both Microsoft 365 Copilot and GitHub Copilot are model-diverse by design without locking customers into a single provider. Different models — like GPT-5.5 or Claude Opus 4.8 — serve distinct roles with different economics. Matching the right intelligence to each task optimizes performance and cost.

Your IQ. Agents struggle with raw data. Significant compute is spent interpreting structure and context before useful work begins. The Microsoft IQ platform empowers your IQ by turning raw data into usable intelligence, continuously building a semantic understanding of how your organization operates across Microsoft 365 and line-of-business systems. It provides agents with the context they need upfront rather than requiring them to reconstruct it. The result is measurable: faster execution, higher accuracy and lower token usage. This is how intelligence compounds within your organization.

Financial operations. FinOps became critical when companies moved to the cloud and requires even greater attention as AI shifts from fixed pricing to usage-driven models. With Foundry and Agent 365, we are providing tools to help our customers optimize their AI costs today.

Frontier business models

Business models are evolving as we use AI to drive business outcomes. The User Subscription License (USL) has become the foundation, providing a package of capabilities for a predictable per-user-per-month fee. Usage-based licensing has emerged for long-running, multi-tasking agents, where cost aligns directly to the work performed.

Microsoft gives customers a unique combination of business model flexibility and integrated product experiences that is unmatched in the market. Microsoft 365 Copilot and GitHub Copilot use both models — a USL offering with not only value and capabilities, but flexible consumption on top. Today we’re announcing the general availability of Copilot Cowork worldwide, which requires the Microsoft 365 Copilot USL and is then usage-based.

Our model-diverse strategy allows customers to purchase capacity with the flexibility to use the right model for the job based on model strengths, economics and the latest innovations. Microsoft Agent Factory provides a single consumption model spanning Microsoft 365 Copilot (including Cowork), GitHub Copilot and agents built in Fabric, Foundry and Copilot Studio.

Our integrated product experiences put AI in the flow of work for both knowledge workers and software developers and manage capacity fluidly across the two. Historically these personas have been distinct, but increasingly the line between them is blurring. Coding is becoming a mainstream knowledge worker skill and chat and Cowork are becoming modalities important for software development. With Microsoft 365 and GitHub, we offer market-leading tools for both roles and make it easy to seamlessly manage capacity based on availability and need.

Agent 365: The control plane

As organizations adopt agents from Microsoft, another provider or build their own, a control plane is essential. Agent 365 gives IT and security leaders a single place to observe, govern, manage and secure agents across the organization. It builds on the Microsoft stack that enterprises trust: Entra for identity, Defender for threat protection, Purview for data governance and Intune for endpoint management. We are extending Agent 365 to include cost management, so organizations can monitor and manage agent spend alongside security and compliance. As the Frontier Firm operating model takes hold, leaders will manage human and agentic work as a single system, with visibility into both performance and cost.

The two most essential elements in any AI solution are Intelligence + Trust. At Microsoft, this conviction shapes how we design every layer of our AI platform. Microsoft IQ enables organizations to harness their own unique IQ, bringing context to data and embedding AI directly into the flow of work to deliver faster, more accurate and more trusted outcomes while safeguarding assets and protecting intellectual property. Agent 365 provides that trust layer, ensuring every agent and AI artifact is observed across the environment so organizations can move decisively from experimentation to enterprise impact with confidence. As Jay Parikh put it at Build, AI alone will not change your business. The system running it will.

We have built this system for our customers and partners, where intelligence compounds from within and every agent operates with control, visibility and trust. Together, we can scale human ambition and define how AI delivers measurable business impact across every role, organization and industry.

Judson Althoff is the chief executive officer of the commercial business at Microsoft. He is responsible for the product strategy, sales, services, support, marketing, operations and revenue growth of the company’s commercial business, which operates in more than 120 regional and national subsidiaries globally.

The post Achieving success with AI appeared first on The Microsoft Cloud Blog.

This momentum is changing not only how insights are generated, but also how decisions are prepared, governed, and executed. As Microsoft Build 2026 made clear, the next phase of agentic AI is not just better outputs. It is governed execution grounded in enterprise context, with systems designed to coordinate action across research, portfolio construction, risk, and compliance under human supervision. For asset managers, this shift is profound.

This is not simply a technology transition; it is a fiduciary moment. For asset managers, Microsoft’s Frontier Tuning points toward institution-shaped intelligence: models tuned not only on enterprise knowledge, but on the workflows, conventions, and decision patterns that define how a firm operates inside its compliance boundary. In a fiduciary setting, that is indispensable because trust depends not just on what AI knows, but on whether its outputs and actions align with approved controls, human oversight, and reproducible governance.

The critical challenge, then, is to maintain trust. AI cannot merely improve performance; its actions must be governed, explainable, and attributable in line with fiduciary duty.

From analytics to homeostasis: AI as a trust-control layer

In biology, homeostasis is how living systems stay stable by sensing change, making small corrections, and restoring balance. In asset management, agentic AI makes it possible to build a similar closed-loop layer across workflows defined by governance parameters, helping keep activity within agreed boundaries of risk appetite, policy, and intent even when markets move faster than humans can triage.

Historically, AI in asset management has centered on prediction, such as forecasting spreads, detecting anomalies, or scoring counterparties. Agentic systems add the ability to observe, decide, and act across workflows under human guidance, gathering evidence, reconciling sources, drafting outputs, escalating exceptions, and coordinating next steps. Used well, these systems can create a homeostatic effect, helping firms maintain operational equilibrium and resilience.

The shift to a homeostatic approach

AI is being used in capital markets not only to deliver new capabilities but to help strengthen trust as a cornerstone of infrastructure. Trust can be infused throughout the platform by design, helping to ensure that data and operations remain safe, responsive, and governed even in fast-changing conditions.

In trading workflows, for example, the LSEG and Microsoft partnership is integrating licensed market data into Microsoft 365 Copilot and agent experiences so that decisions can be informed with authoritative, permissioned context at the point of execution. This is increasingly the right pattern for agentic systems: not intelligence separated from governance, but intelligence grounded in trusted context and controlled inside the systems where work happens.

By combining licensed data with built-in governance capabilities, trust is embedded into how intelligence is delivered, orchestrated, and acted on across the workflow, rather than managed downstream.

Likewise, Moody’s has federated decision-grade credit intelligence directly into Microsoft 365 Copilot, Researcher, and Excel, grounding AI-assisted analysis in authoritative, auditable context so that trust can be enforced by the system, not retrofitted after the fact.² Similarly, Morningstar has embedded proprietary research as entitlement-aware context for copilots and agents, keeping advisor workflows reviewable and aligned to licensed sources and user permissions. UBS is also bringing together internal and market data into AI-assisted advisor workflows, enabling client advisors to access unified, permissioned insights in real time.

Further, at Nasdaq, AI is being applied directly into the boardroom experience while preserving the strict controls those workflows demand. Its Nasdaq Boardvantage platform helps members turn hundreds of pages of board materials into concise, decision‑ready insights. This reduces review time by up to 60% while ensuring that outputs are fully auditable, remain grounded in proprietary data, and are protected by design.

Together, these examples show that intelligence can scale when workflows are auditable, permissioned, and attributable by default.

Getting started: Three moves asset managers can make now

The path forward is not theoretical; it is operational. Asset management firms can begin by focusing on three priorities:

• Invest in data readiness. Start by unifying internal and external datasets into a governed data foundation that can support agentic AI. Platforms like Microsoft Fabric are increasingly designed not only to bring data together, but to create the shared business context that agents need to operate consistently across teams and workflows. That matters because the bottleneck in scaling agents is no longer just model capability. It is whether each system can access the same governed definitions, permissions, and operational data foundation needed to move from isolated experiments to production-ready agent systems.

• Operationalize governance and enterprise ontology. Define ownership, approval paths, and audit checkpoints directly into AI-enhanced workflows under human supervision. A broader contextual intelligence layer is now emerging to connect how people work, how the business is modeled, and how knowledge is grounded across enterprise systems. Within that picture, Microsoft Work IQ remains an important component, helping make sense of activity across emails, meetings, documents, and chats while honoring existing permissions, sensitivity labels, and governance controls. Applied consistently in AI-assisted experiences, this kind of shared context helps turn governance from a static framework into a living system embedded in everyday work.

• Enable model-agnostic intelligence. Explore the comprehensive catalog of AI models so that you can select the best ones as needs evolve. Microsoft Foundry increasingly represents the production layer for agents, bringing together models, hosted agent services, memory, observability, evaluation, and guardrails in one governed environment. That gives firms more flexibility as requirements change while helping them move from pilot use cases to durable, production-scale systems.

Once these foundations are in place, firms can layer agentic workflows across research, operations, and risk while keeping clear human ownership at every material decision point and standardizing evidence trails so material outputs are reproducible and traceable. The goal is not automation alone, but structural operating leverage that scales intelligence with observability, governance, and security.

Making trust measurable and operational

Trust is not abstract in capital markets; it is the product of repeatable controls including lineage, entitlements, validation, and oversight. In practice, this is increasingly operationalized within the systems where work already happens. The idea is to bring productivity, security, identity, and agent management into a more unified operational layer, so that trust-related controls are enforced upstream in the flow of work rather than applied after the fact. Microsoft 365 E7 for Enterprise is one example of this broader architectural direction.

Accordingly, many firms focus on AI-enabled workflows defined by a common set of “trust signals,” including:

• Traceability: Every output must link back to authorized data, policies, and model versions.
• Explainability: Decisions must be not just technically correct but contextually defensible.
• Human accountability: Ownership must be explicit: who reviewed, who approved, and who can intervene.
• Controls and monitoring: Guardrails must operate continuously, especially under stress conditions.
• Regulatory evidence: Every decision path must be reproducible when challenged.

These are not abstract ideals but key operational expressions of responsible AI practices, helping asset managers ensure that AI is compatible with fiduciary duty.

The bottom line: Improving operational leverage

As firms mature in AI adoption, the focus evolves from products to platforms and ultimately to systems. The difference is not incremental; it is structural.

Value is no longer created by isolated use cases, but by integrated intelligence loops running on trusted infrastructure that connect external market signals with internal data, workflows, and decision rights. When orchestrated well, AI can serve as a coordination layer for action at scale in line with fiduciary requirements.

That is what drives operational leverage. Instead of adding incremental tools or headcount, firms can embed intelligence directly into workflows so the organization can better align decision-making, sensing, and action. Trust is thus inherent within systems and is no longer merely a byproduct.

Take the next step

We work with institutions to move from experimentation to production by embedding governance into multi-agent systems and operations so that fairness, reliability, privacy, and accountability are built into the foundation where AI runs. Our end-to-end approach can help unify enterprise and market intelligence into a shared operating context that turns fragmented insights into coordinated action at scale. Because most financial markets professionals already use Microsoft tools, it can also help lower deployment effort and improve adoption while keeping human oversight at every material decision point.

This is a rare and urgent invitation to the custodians of fiduciary capital to architect a homeostatic system of intelligence. It marks the convergence of capital and cognition, redefining how enduring organizations create and sustain value over time.

Explore how to best adopt AI

To learn more about how your organization can adopt AI, start by engaging with your Microsoft representative or service provider and explore these resources:

• Read our blog for stories about how Frontier Firms are using agentic AI to gain an edge in capital markets.
• To learn how Frontier Firms in capital markets lead with AI, visit Microsoft for Financial Services.
• For a practical transformation framework, read our new ebook, Frontier Transformation in Capital Markets.

¹ EY, Unlocking strategic advantage: Generative AI in wealth and asset management, September 16, 2025.

² Moody’s, Moody’s advances decision-grade credit intelligence across enterprise AI workflows, powered by Microsoft 365 Copilot, April 21, 2026.

The post Trust as infrastructure: How agentic AI is rearchitecting asset management at scale appeared first on The Microsoft Cloud Blog.

Building on that foundation, a growing community of customers and partners are fueling rapid innovation by fine-tuning new models, deploying AI applications, and developing specialized agents that expand what’s possible across the diagnostic imaging ecosystem. This momentum is shaping the next era of radiology—one defined by continuous innovation, open collaboration, and powerful new ways to connect insights from image to action.

Listening first: How customer feedback shapes every innovation

For decades, PowerScribe has been built alongside radiologists, grounded in real-world workflows and shaped by continuous feedback and close clinical partnerships with healthcare organizations across the country. This approach, building with radiologists and grounding innovation in real-world use, is fundamental to how we design and evolve our solutions, especially when it comes to performant AI. Those insights directly shaped how we evolved to PowerScribe One, where preserving the workflows and integrations that teams rely on while introducing a more modern, cloud-enabled experience designed for what comes next.

We’ve invested in dedicated voice-of-customer programs and teams whose sole focus is to continuously gather feedback. From advisory boards, clinical partnerships, and real-world usage, we translate those insights directly into our roadmap. This isn’t a one-time input; it’s an ongoing loop that ensures the capabilities we deliver reflect the evolving needs of radiologists across a wide range of environments.

That’s why we partner closely with organizations like University of Rochester Medical Center (URMC), St. Luke’s University Health Network (St. Luke’s), along with many others, through early preview programs ahead of general availability, so they can guide how innovation needs to be integrated. By embedding structured preview and validation stages into our development cycle, we align our releases with customer readiness, continuously refining based on real-world feedback. The result: technology that not only pushes boundaries, but prioritizes the workflow and overall customer experience.

Ultimately, it’s this approach, continuous collaboration grounded in the day-to-day realities of radiology, that gives us confidence in how we are shaping the future of the reporting workflow. This foundation makes these customer stories not just possible, but repeatable at scale.

PowerScribe One serves as the foundation for what’s next

At URMC and St. Luke’s, trust in PowerScribe One began with confidence in a cloud-based foundation designed to scale and integrate seamlessly into the radiologist’s workflow. For URMC, moving to the cloud was essential to unlock advanced AI capabilities that improve efficiency and provider satisfaction amid rising volumes and increasing cognitive demands. At St. Luke’s, modernization with cloud capabilities was equally strategic, enabling innovation while maintaining continuity and trust across the enterprise.

Our partnership and deep engagement model with URMC and St. Luke’s are reinforced at scale: today, more than 10,000 radiologists across 250+ organizations have migrated to PowerScribe One, generating millions of reports every week, across environments ranging from large Integrated Delivery Networks (IDNs) and academic medical centers to independent reading groups. That experience shaped a clear understanding of how to bring AI into the reporting workflow—not as a separate tool, but as a capability embedded directly where radiologists work, without introducing additional steps or fragmentation.

Both organizations are realizing real outcomes through PowerScribe One and its AI features, including generated draft impressions personalized to each radiologist that support improved efficiency and report quality.

“We chose PowerScribe One so we could really take advantage of cloud-based reporting. It gives our radiologists built‑in AI, excellent speech recognition and personalized impressions, making it easier to keep up with increasing demands while continuing to deliver great patient care. Microsoft has been with us every step of the way, staying responsive and supportive through implementation, go‑live and ongoing adoption. We will continue this partnership to continue to improve our workflows and efficiency.”

Robert Fournier, MD, Chairman of Radiology, St. Luke’s University Health Network 

At URMC and St. Luke’s, generated draft impressions were widely adopted because the feature works natively inside the reporting workflow—helping ensure key findings are pulled from the report and summarized in the impression section, reinforcing radiologists’ confidence in their report quality.

The ongoing adoption of PowerScribe One and its draft impression capabilities reflects a broader principle: when AI is fully integrated into the workflow, it enables radiologists to deliver more consistent, efficient, and high-quality reports without disrupting how they work.

Extending AI in the reporting workflow with Dragon Copilot

Now, URMC and St. Luke’s are extending these capabilities with Dragon Copilot, building on PowerScribe One to introduce intelligent summarization and automation directly within the reporting experience. Both organizations are actively leveraging prior report summarization, a feature within Dragon Copilot, to surface essential patient context from relevant prior reports, helping radiologists interpret studies with greater clarity and focus. At URMC, this capability is already delivering value by improving visibility into patient history.

“It works amazingly…it provides a great interface for seeing so much about the patient you otherwise might not see.”

Sean Cleary, MD, Vice Chair of Informatics for Imaging Sciences, University of Rochester Medical Center

Looking ahead, both organizations see significant potential as Dragon Copilot continues to evolve. As it gains access to richer patient context and connects to a broader ecosystem of first- and third-party AI applications and agents, Dragon Copilot can help to further reduce cognitive load and enable continuous innovation without disrupting the radiologist’s workflow.

Meeting customers where they are: From deploying off-the-shelf AI to fine-tuning models

Increasingly, innovation in radiology is shaped not just by what Microsoft delivers, but by how customers and partners extend AI within real-world workflows—helping radiologists work more efficiently, surface critical insights faster, and support better patient care.

As AI adoption expands across radiology, organizations aren’t moving along a single path; they’re navigating a wide range of needs simultaneously. Some are focused on deploying trusted, ready-to-use AI solutions directly into clinical workflows, while others are exploring how to build, customize, and push the boundaries of what’s possible with AI. At Microsoft, we’re designing with this range in mind to meet customers where they are and support multiple approaches to innovation.

For organizations looking to quickly operationalize AI, we provide a streamlined path forward with centralized access to a curated set of FDA-cleared third-party imaging AI applications from our ecosystem of partners—helping simplify how they are evaluated, deployed, and integrated. These applications integrate with our reporting workflows, enabling radiologists to access AI-powered insights within PowerScribe One and helping simplify the adoption of new capabilities.

For St. Luke’s, this approach enabled the rapid deployment of a fracture detection model from Gleamer, delivering immediate impact across its geographically distributed network and helping ensure more consistent diagnostic support regardless of where patients entered the system.

In addition to bringing FDA-cleared imaging AI into practice today, we provide the flexibility for customers and partners to build, customize, and extend AI capabilities as their needs evolve. Our premium medical imaging foundation models, MedImageInsight Premium and CXRReportGen Premium, can be requested for preview through Microsoft Foundry, and are designed for fine-tuning across modalities and workflows. These models are not medical devices, but they enable teams to build and fine-tune models that can complement clinically validated imaging AI solutions.

Delivered as fully managed endpoints, our premium models are continuously improved with curated data and enable AI builders, health systems, and partners to develop institution-specific solutions tailored to local data, specialty use cases, and evolving clinical needs. Models derived from CXRReportGen Premium can be integrated into experiences like Dragon Copilot, bringing high-performing AI directly into the radiologist’s workflow for summarization and report generation.

Together, this approach allows organizations to combine production-grade, regulated AI with ongoing innovation on a single platform, bridging standardized diagnostics and bespoke AI development. Companies like Milvue, a radiology-focused AI developer, are already using our models to accelerate development of solutions tailored to real-world clinical workflows.

“Milvue is building a radiology-native VLM. By working with Microsoft and leveraging CXRReportGen, we could start from a strong foundation allowing our team to focus on what matters most: turning foundation-model capability into clinically validated, workflow-ready radiology solutions.”

Alexandre Parpaleix Co-Founder/CEO, Milvue

No matter where customers and partners are in their journey with generative, multimodal, and agentic AI, we’re here to support them. From clinical applications like PowerScribe One and Dragon Copilot to customizable models from Microsoft Foundry, we provide a trusted, scalable foundation for innovation—enabling organizations to advance at their own pace while keeping workflows, performance, and outcomes at the center.

We’re excited to bring this next wave of radiology innovation to life at the SIIM26 Annual Meeting + InformaticsTECH Expo. Join us in Pittsburgh, PA to experience it firsthand. Visit us at the SIIM 2026 Booth #630–632 where customers and partners can explore our solutions, see live demos, and engage with our models in an interactive learning lab. See what’s possible when AI is truly embedded in the workflow.

See how AI fits into your radiology workflow

• Connect with our team
• Read more about migrating to PowerScribe One
• See demo of how Dragon Copilot supports radiologists and other members of the care team
• Learn how Microsoft applies AI for better health outcomes

The post How collaboration advances workflow-native AI appeared first on The Microsoft Cloud Blog.

As Australia looked to extend its winning streak to four straight Ashes on home soil, it was clear that left-handed batter Travis Head was leading the way for Australia as the runs piled up. But just how good was his performance? Fans using the Cricket Australia Live app had an instant answer. 

Thanks to the app’s new AI Insights feature, which provides live insights on player milestones, records and key moments using OpenAI’s GPT-5 within Microsoft Foundry, cricket aficionados and newcomers can now access much-needed context to better engage with the game. They can also dig deeper by asking follow-up questions about the insights provided. It’s an exciting development for Cricket Australia, the governing body of the sport in the country. 

“The recent series where England were here in Australia had a couple of key moments where I saw the insights come to life in real-time,” says Cricket Australia CEO Todd Greenberg. “And you can see the engagement through the analytics and the tracking that when something is delivered in the right time frame, in the right format, into the right hands, it has a huge effect.” 

Indeed, AI Insights showed that Head’s 172 runs for the match were his fifth-highest aggregate total in a test. His only higher efforts were 220 runs against Sri Lanka, 213 against West Indies, 181 against India and 180 against England. Head’s big day earned him Player of the Match honors and helped Australia claim a five-wicket victory in the match and a 4-1 Ashes series victory against its archrivals. 

Going beyond the box score 

“Scores and highlights tell you what happened. But the context tells you why you should care about it,” says Balamurugan P M, chief technology and digital officer at Cricket Australia. 

“It comes down to the storytelling. From my perspective, I thought it was essential for fans to learn more about the story rather than just following the scores or watching highlights. So, we wanted to give a different experience.” 

Cricket Australia had a corker in its arsenal as AI Insights came into focus – an extensive archive of official scorecards that dates to 1886, providing a wealth of historical data that could bridge the gap between the past and present. Those scorecards were carefully integrated over a period of three months to ensure the information would pass muster among the serious cricket experts. 

“We had hundreds of years of data, and when it comes to fans, trust is non-negotiable,” Balamurugan says. “When you’re dealing with records and milestones, you can’t make mistakes. There are some hardcore fans who know these stats like the back of their hand. History is core to cricket’s identity. And instant context turns a scoreboard into a story. 

“Getting that volume of data, integrating it and surfacing greater context for live games required huge data alignment and validation. With our systems and with the skilled team that we’ve got, that was made possible.” 

Creating a solution fans can use in real time 

Cricket Australia joined forces with Microsoft, alongside technical partners Insight Enterprises, HCL Tech and Skewer, to create the new iteration of the app. With the important Ashes and T20 international tournaments on the horizon, time was of the essence to launch the app before the bats were raised on those key fixtures. 

The app is anchored by Microsoft Azure, the cloud foundation that Cricket Australia uses to run and scale its digital platforms and the app experience. AI Insights takes advantage of Azure OpenAI Service in Microsoft Foundry, which generates the real-time, match-aware insights that serve as a companion to what fans are seeing on the field. 

“What we’re talking about is a really good example of solving a fan-facing problem with deep technical capability and a shared vision on delivery,” Greenberg says. “Microsoft brought world-class cloud and AI foundations. Without them, we would not have been able to get as far as we have. And our partners have helped accelerate the build, the integration and, importantly, operational readiness.” 

One of the biggest challenges with AI Insights is ensuring that fans watching a match and using the app can get updates and context within the flow of the game, making it an additional resource for fans at the grounds or watching alongside with commentary. 

Azure Cosmos DB supports Cricket Australia’s ecosystem of apps – including Cricket Australia Live with AI Insights and PlayCricket, which hosts scores for up to 7,000 community matches a weekend. The technology provides a fast, scalable data layer that can update quickly during live play, always keeping fans aware of the latest scores. 

“All live sport has one thing in common. There are no pauses,” Greenberg says. “It’s not like reality television. So, the experience has to be fast, reliable and consistent, especially when it’s under peak demand and when you have millions of people enjoying it at the same time.” 

An experience for every type of fan 

While cricket has its ardent supporters, especially in Australia, it can also be difficult for newcomers to pick up. As Cricket Australia looks to cultivate the next generation of fans, Greenberg realizes that the app can prevent sticky wickets for the sport’s novices. 

“I mean, we play a crazy sport that goes over five days and sometimes at the end of the five days, you still don’t get a result,” Greenberg says. “We can’t expect people to be tuned in at every moment, but what we can do is we can hyper-personalize the way they would like to engage with the sport during the contest.” 

The Seddon Cricket Club in Melbourne has been in existence since the 1920s and is now home to several senior, junior and all abilities sides that compete in associations across Australia. It is also home to a loyal supporters group, featuring fans who love the game in all forms. For them, the AI Insights on the Cricket Live App has been a value add as they go deeper into the game. 

“It’s definitely made it more interesting to follow along and learn more about the players,” says Cassie Gray, a Seddon Club supporter and cricket fan. “You could follow a player, you could see what they’re known for, as well as figure out what’s their next step or what do they need to get an amazing moment next. 

“Cricket is a game of history. It’s been around for a really long time, and the players influence other players, and countries influence other countries. With the insights, it gives me an understanding of not just what’s happening today, but what’s led up to that in the game itself.” 

The next step for AI insights is to create greater personalization within its levels of information for different types of fans. A user can select “newcomer,” “history buff” or “stats guru” and receive insights tailored to their persona. 

“We want to understand every fan and cater to how they want to be served by the app,” Balamurugan says. “We have moved from scores to storytelling, but we want to move from storytelling to fans setting up the narrative themselves. Fans should hear the story how they want to hear it. That is one of our lodestars.” 

With the initial success of the AI Insights feature, Greenberg said other sports organizations have reached out to learn more about how it was developed and the impact on the fanbase. Most people working at Cricket Australia have a deep love of the sport, often having played for many years. Greenberg hopes the app’s success and further innovation can continue the sport’s momentum. 

“The thing we’ll never know until much later on is the impact that we’re having on young kids falling in love and choosing cricket as their preferred sport,” he says. “And if we help them love it, what we can create for a fan on their journey between the ages of 8 and 80 is astronomical for a sport like cricket. And so, we’re very mindful of ensuring kids get the opportunity to engage in cricket so we can form lifelong partnerships.” 

Top Image caption: Supporters at the Seddon Cricket Club in Melbourne love the game in all forms, and the Cricket Live App featuring AI Insights has allowed them to gain further insights into the sport, whether they are a novice fan or stats guru. Photo by Graham Denholm for Microsoft.  

Elliott Smith writes about AI and innovation at Microsoft, from how the Premier League is transforming its online presence to why AI may play a major role in saving the Amazon rainforest. Previously, Smith worked as a sports reporter in Washington, D.C., Washington state and Texas, covering high schools to the pros. You can contact him on LinkedIn. 

The post Cricket Australia uses AI Insights to bring fans closer to the action appeared first on The Microsoft Cloud Blog.