In production AI systems, different problems require different model capabilities. Document ingestion, reasoning, coding, and automation workflows rarely rely on a single model. They depend on select...Update Type: Announcement, Services: Microsoft Foundry, Categories:
updateMicrosoft Sentinel platform offers a growing list of tools and features, with graph being a cornerstone capability. Sentinel graph is a relationship-first method for organizing and querying data wi...Update Type: Announcement, Services: Microsoft Sentinel, Categories:
updateAzure NetApp Files migration assistant (with SnapMirror) provides efficient and cost-effective data migration leveraging ONTAP's built-in replication engine for seamless transition from on-premises or CVO/other cloud providers to Azure NetApp Files (ANF).
previewThe Model Context Protocol exploded onto the scene because it's easy. Stand up a server, expose a few tools, point Claude or VS Code at it, and your agent can suddenly read files, hit APIs, and run c...Update Type: Announcement, Services: App Service, Categories:
updateA business-first approach to economics on Microsoft FabricEnterprise planning is evolving across industries - from an isolated finance exercise to a cross-functional capability that spans finance, operations, supply chain, HR, and executive leadership, but the tools and pricing models behind it have not kept up. Planning in Microsoft Fabric (Preview) is bringing planning, analytics, reporting, and data management into one integrated experience in Microsoft Fabric, moving organizations from disconnected processes to real-time, data-driven decisions.As customers adopt Fabric Planning, one question comes up: How does billing work and why is it designed this way? This blog post explains how the model is built for outcomes, impact, and predictability, not just usage.Why pricing for Fabric Planning is fundamentally different from traditional EPM solutionsEnterprise planning behaves differently from traditional analytics or data workloads. Analytics and data workloads are continuous, predictable, and incremental. Planning is cyclical with high-intensity usage during month-end, quarterly forecasting, and annual planning and lower levels of interaction between cycles. This breaks the two common pricing models:Pure consumption pricing undervalues planning, because the critical work happens in short bursts.Seat-based licensing feels rigid, forcing you to pay for users who participate only occasionally.Fabric Planning introduces a hybrid pricing model that reflects real-world usage patterns, with role and session-based pricing for users and job-based pricing for automation. It’s predictable during peak cycles and flexible across a broad set of participants.Figure: User roles segregated with different capabilities.Role-based pricingEach role contributes differently to the planning lifecycle, and the pricing reflects that:Viewers: Executives, decision-makers and business, who consume plans and insights. Read-only access; priced low to drive adoption.Stakeholders: departUpdate Typ
previewMulti Cloud Data integration and transformation made easy and painless by Fabric Data Factory.Update Type: GA, Services: Microsoft Fabric, Data Factory, Categories:
releaseA couple of months ago I wrote about scaling MCP servers behind App Service's built-in load balancer. The trick back then was to lean on stateless HTTP transport so any instance could serve any reque...Update Type: Announcement, Services: App Service, Categories:
updateComplex cloud environments have outpaced manual operations. Agentic cloud operations connect people, tools, and data to streamline investigation workflows and move teams from scattered signals to evi...Update Type: GA, Services: , Categories:
previewOne of the things that makes agents genuinely useful is that they can remember. Instead of treating every interaction as completely new, an agent can carry context across sessions, recall preferences...Update Type: Announcement, Services: Microsoft Foundry, Categories:
securityWhat if your cloud environment could help you move from insight to action in real time, with systems already working through the next set of decisions?Update Type: Announcement, Services: Azure AI Services, Categories:
updateAzure SDK releases every month. In this post, you'll find this month's highlights and release notes. The post Azure SDK Release (May 2026) appeared first on Azure SDK Blog.Update Type: Announcement, Services: , Categories:
updateModel availability in Microsoft Foundry is region-dependent. The region approved for your project may not be the one where the model or Foundry Agent Service support you need is available. That creat...Update Type: Announcement, Services: Microsoft Foundry, Categories:
updateAutoJack is a novel exploit chain showing how a single malicious webpage can turn an AI browsing agent into a remote code execution vector on the host machine. By abusing trust in localhost, missing authentication, and unsafe parameter handling, attackers can trigger arbitrary process execution through AutoGen Studio’s MCP WebSocket. The research highlights a broader pattern - when agents can browse untrusted content and access local services, traditional boundaries like localhost are no longer secure. The post AutoJack: How a single page can RCE the host running your AI agent appeared first on Microsoft Security Blog.
securityThe new preview capabilities in Fabric Data Warehouse for approximate string matching, along with modern string-processing functions and operators simplify everyday string processing with T‑SQL language. Together, these additions help developers handle variation directly in SQL while improving query clarity and portability.Update Type: Announcement, Services: Microsoft Fabric, Categories:
previewIntroduction Generative AI (GenAI) is poised to transform the construction industry by addressing chronic challenges such as low productivity, cost overruns, schedule delays, and labor shortages. B...Update Type: Announcement, Services: Azure AI Services, Categories:
updateWe shipped a lot at Build 2026: hosted agents, Toolboxes, Foundry IQ, Memory, Managed Compute, fine‑tuning, Frontier Tuning, and a new evaluation and optimization stack. Read as a feature list, it is a lot to hold in your head. So here is a simpler way to see it: these are the parts you need to […] The post Outcome-driven learning systems: Enterprise RL with OpenEnv and Foundry appeared first on Microsoft Foundry Blog.Update Type: Announcement, Services: Microsoft Foundry, Categories:
updateThe problem we're solving Previously, Microsoft Entra identities in Azure SQL Database could only be created as contained database users - principals scoped to a single database with no server-leve...Update Type: GA, Services: SQL Database, Categories:
releaseSummer heat meets the world stageThe FIFA World Cup 2026 kicked off this month across 16 host cities in the United States, Mexico, and Canada—where it is currently summer, including in cities like Miami, Dallas, Houston, and Atlanta. These cities aren’t exactly known for mild June weather. Player safety protocols, fan comfort advisories, and broadcast scheduling all depend on real-time environmental conditions at each venue. The data exists (weather APIs publish readings every few minutes), but standing up parallel monitoring infrastructure for 11 cities?That’s traditionally a full afternoon’s worth of infrastructure setup, portal configurations and query and dashboard authoring.We built a real-time weather monitoring system for all 11 US-based World Cup venues in under five minutes using natural language prompts and Fabric Eventstream and Eventhouse AI Skills. No portal clicking. No copy-paste-modify-repeat.From 42+ steps to one sentenceBuilding a single Eventstream pipeline through the Fabric portal requires navigating menus, selecting source types, configuring properties, wiring operators, choosing destinations, and publishing. That’s roughly 42 interactions (clicks and text box entries) and more than five minutes per source. Multiply that by 11 cities and you’re looking at 60 minutes or more of repetitive work before any data starts flowing.With Eventstream AI Skills for Fabric, all we needed to do is type a single prompt describing the desired outcome:Create an eventstream named WorldCupUSCitiesWeatherFeed for all 11 US World Cup host cities including Miami, Dallas, Atlanta, Houston, Seattle, Los Angeles, Philadelphia, Kansas City, San Francisco, New York, and Boston. Ingest real-time weather for each city, filter for heat-stress conditions where relativeHumidity exceeds 70 percent, and land the filtered events in my WorldCupCities KQL database. Set the optional locationName field in the weather feed to the name of the city the weather feed is for.The AI skillU
updateIntroduction A common pattern in large Azure deployments is to route VNet-to-VNet traffic through Microsoft Enterprise Edge (MSEE) routers. This happens when spoke VNets in a hub-and-spoke topology...Update Type: Announcement, Services: Virtual Network, Categories:
updateNew Forrester Total Economic Impact™ study shows Microsoft Security consolidation delivers ROI, lowers risk, and prepares organizations to secure AI. The post New Forrester study shows customers who unified with Microsoft Security benefited from 124% ROI appeared first on Microsoft Security Blog.
securityUpdates, releases, and news
