Microsoft Build 2026: Windows, AI Models, Scout Agent, and Project Solara — Everything Announced At The Conference

 

Microsoft kicked off its annual developer conference, Build 2026, with the announcement of a broad set of updates spanning Windows, artificial intelligence (AI) models, hardware, and experimental platforms. Instead of centring on a single product, the announcements collectively signal a transition in which AI moves from a feature to a foundational layer across the Windows ecosystem.

 

A substantial portion of the announcements addressed Windows itself, which Microsoft is increasingly positioning as a platform for AI workloads rather than purely a desktop operating system. New features include support for Linux containers through Windows Subsystem for Linux, updated configuration tools for development environments, and an Intelligent Terminal that integrates AI assistance directly into command-line workflows.

Microsoft also expanded its Windows AI APIs beyond devices with dedicated AI hardware, enabling a broader range of PCs to run AI-powered features such as speech recognition and video enhancement locally.

To support the deployment of AI agents on Windows, Microsoft outlined a set of security capabilities designed to ensure these systems operate within defined limits. These include agent containment through execution containers, identity management for AI agents, integration with Microsoft Entra and Intune, security protections through Microsoft Defender and Purview, and support for Windows 365-based agent environments.

 

Microsoft introduced Scout, an always-on AI agent integrated across Microsoft 365 services, including Teams, Outlook, OneDrive, and SharePoint. Unlike traditional assistants that require explicit prompts, Scout is designed to operate continuously in the background, drawing on emails, calendars, chats, and documents to understand user context.

According to Microsoft, Scout can help prepare for meetings, manage scheduling conflicts, draft emails, and surface relevant information without manual input. The company positions Scout within a new category of AI systems it calls "autopilots", designed to execute tasks on behalf of users rather than simply respond to queries.

 

Microsoft expanded its in-house AI capabilities with seven new first-party models under its Microsoft AI (MAI) family, available through Microsoft Foundry. These AI models take text, code, image, voice, and speech as inputs.

The lineup includes MAI Image-2.5 and a faster Flash variant for image generation, MAI Transcribe-1.5 for speech-to-text, MAI Thinking-1 for multi-step reasoning, MAI Voice-2 and its lower-latency Flash version for text-to-speech, and MAI Code-1 Flash for code generation and completion.

The additions extend Microsoft's ability to offer a full-stack AI ecosystem across multiple modalities without relying entirely on third-party models.

 

Alongside the MAI family, Microsoft introduced the Aion model lineup — a set of small language models designed to handle AI workloads locally on Windows systems, reducing dependence on cloud processing.

range includes Aion 1.0 Instruct, built for Windows 11 to support on-device tasks such as text summarisation, rewriting, and accessibility features, and Aion 1.0 Plan, a 14-billion-parameter reasoning and tool-calling model capable of supporting complex, agent-driven workflows even in offline environments. Microsoft said these on-device capabilities address latency, privacy, and cost constraints associated with cloud-based processing.

 

Microsoft used Build to present its longer-term vision through Project Solara, a platform designed for what the company calls "agent-first devices." Built around AI agents that interpret user intent and execute tasks across services, Solara replaces the conventional app-based model with background orchestration of workflows.

The platform introduces a "just-in-time UI" concept, where interfaces are generated dynamically based on context rather than being pre-designed for specific devices. Microsoft demonstrated the approach through prototype hardware, including a wearable badge and a desk-based system, both of which rely entirely on agents rather than traditional applications.

 

To support the broader shift towards on-device AI processing, Microsoft also unveiled new hardware platforms. The Surface RTX Spark Dev Box is a compact system designed to handle AI workloads locally. It is powered by Nvidia’s newly launched RTX Spark chipset, which was launched in the ongoing Computex 2026 event in Taiwan.

The DGX Station for Windows, developed in partnership with NVIDIA, is built to run large AI models, including those with hundreds of billions of parameters.

Both systems reflect a push to bring more AI processing closer to the user, reducing reliance on remote data centres.

 

Microsoft also shared progress on the quantum computing front with the Majorana 2 chip. It introduced a revised materials approach that, according to the company, significantly improves qubit reliability compared to its predecessor. Microsoft said the development could accelerate its roadmap towards building a scalable quantum computer, with a target timeline of 2029.

The company also noted that its AI systems played a role in developing the chip, highlighting how machine learning is being applied not just in software, but in scientific research and hardware design as well.