Microsoft has confirmed its annual developer conference, Build 2026, will take place on June 2 and 3 at the Fort Mason Center in San Francisco. The move ends a decade-long run in Seattle and signals a deliberate pivot toward a smaller, more focused event centered on AI agents, trust in technology, and a fundamental rethinking of the developer platform.

The choice of Fort Mason Center—a scenic cultural venue with a capacity of roughly 3,000—represents a stark contrast to the sprawling convention halls that have hosted Build since 2017. It marks a return to San Francisco for the first time since Build 2016, when the conference was last held at Moscone Center. Microsoft’s decision to downsize its flagship developer gathering and place it in the heart of the AI boomtown underscores how deeply the company is aligning its entire development story with artificial intelligence.

A homecoming with a purpose

Build’s history with San Francisco runs deep. The conference was held in the city from 2013 through 2016 before moving to Seattle, where it became a massive 5,000-plus attendee event. By choosing Fort Mason, Microsoft is not only embracing a more intimate atmosphere but also inserting itself directly into the Bay Area’s thriving tech ecosystem. San Francisco has become ground zero for AI startups, venture capital, and top-tier engineering talent—assets Microsoft is keen to engage as it competes for the future of AI-driven development.

The downsizing doesn’t imply diminished ambition. Instead, it reflects a curated experience: fewer keynotes, more hands-on workshops, and deeper technical sessions. Attendees can expect a heavy focus on building real-world AI solutions rather than broad-spectrum product announcements. Microsoft’s executive team, including likely appearances from CEO Satya Nadella and CTO Kevin Scott, will use the stage to articulate a vision where AI is not just a feature but the core of the development experience.

AI agents take center stage

The conference’s leading theme—AI agents—points to the next frontier beyond chatbots and copilots. Over the past year, Microsoft has steadily advanced its agentic AI capabilities, from Copilot Studio’s autonomous agents to the experimental AutoGen framework. At Build 2026, developers can anticipate new tooling, APIs, and runtime environments specifically designed for creating, testing, and deploying AI agents that can act independently across Microsoft’s platforms.

These agents are poised to redefine how software interacts with users. Instead of simply responding to commands, they will proactively manage workflows, reason over data from multiple sources, and execute multi-step tasks. Session tracks will likely cover orchestration patterns, memory management, and integration with services like Microsoft 365, Dynamics 365, and Azure. For Windows developers, the implications are profound: native agent APIs could allow desktop applications to spawn autonomous assistants that operate seamlessly within the operating system, blurring the line between user and machine.

Trust as a competitive differentiator

Amid the rapid deployment of AI, trust has emerged as a critical concern. Microsoft is betting that its emphasis on security, compliance, and responsible AI will set it apart from competitors. Build 2026 will dedicate significant stage time to the tools and frameworks that give developers control over AI safety. Expect updates to the Responsible AI Toolbox, new Azure AI Content Safety features, and perhaps a Trustworthy AI certification program aimed at enterprise customers.

Privacy-preserving machine learning, explainability modules, and adversarial robustness testing are also on the agenda. Microsoft knows that for AI agents to become ubiquitous, they must be transparent, auditable, and resistant to manipulation. Workshops on these topics will be directly applicable to Windows development, where local AI models running on Neural Processing Units (NPUs) must still adhere to enterprise security policies. The company’s recent push for confidential computing on Azure and Windows suggests that on-device AI will get similar trust guarantees.

The platform shift: from tools to ecosystems

The third pillar—developer platform shift—signals a broader transformation of Microsoft’s developer story. Gone are the days when Build simply unveiled new versions of Visual Studio or .NET. The platform is now an interconnected web of cloud services, AI models, and runtime hosts spanning the edge, client, and cloud. At Build 2026, Microsoft will chart a course for this reimagined stack.

Azure AI Foundry, the unified platform for building AI applications, is expected to receive major updates. Deeper integrations between GitHub Copilot and Azure services, new multi-agent orchestration tools, and advancements in AI-powered development environments like VS Code will headline the talks. Developers will learn how to combine OpenAI models, open-source models via Azure’s model catalog, and proprietary small language models (SLMs) optimized for Windows devices.

A key component of the shift is the blurring of boundaries between developer and machine. Copilot itself is evolving from an autocomplete tool into a collaborative development partner that can write entire functions, debug complex errors, and even participate in architecture discussions. Build sessions on prompt engineering, fine-tuning, and retrieval-augmented generation (RAG) will be technical deep-dives rather than overviews.

Windows development in an AI-first world

For Windows enthusiasts and developers, Build 2026 will likely be a milestone. Windows is no longer just an operating system for running applications; it is becoming an AI orchestration layer. The past two years have introduced the Copilot key, Windows Copilot Runtime, and AI-powered features like Recall and Click to Do. At Build, Microsoft may reveal how third-party developers can tap into these capabilities through new Windows AI APIs.

Imagine a Windows application that isn’t just a passive tool but an active agent: a spreadsheet that autonomously pulls live data from your email and calendar, generates summaries, and sends reports without explicit instructions. Or a gaming overlay that uses an AI agent to coach players in real time by analyzing gameplay. These scenarios are within reach, and the platform shift Microsoft will detail is designed to make them a reality.

Developers building on WinUI, MAUI, or desktop frameworks like WPF can expect guidance on incorporating AI agents into their apps. The Windows Developer Platform team is likely to showcase hybrid architectures where some AI processing occurs on-device (via the NPU) for speed and privacy, while more complex reasoning offloads to Azure. This distributed intelligence model will be a cornerstone of the “AI PC” narrative that Microsoft has been pushing with hardware partners.

A condensed schedule, not a reduced ambition

With only two days of programming, Build 2026 will be a sprint. The agenda will likely combine an opening keynote with a series of technical breakouts, fireside chats, and a showcase floor. Fort Mason’s waterfront setting, complete with views of the Golden Gate Bridge, underscores the conference’s new character: tightly focused, technically dense, and designed to foster collaboration rather than spectacle.

Registration details, virtual attendance options, and pricing are expected to be announced in the coming months. Historically, Build sells out quickly, and the smaller venue suggests that on-site seats will be even more competitive. Microsoft may offer a digital-only pass as it has in the past, but the emphasis on hands-on labs and networking signals that the live experience will be the primary draw.

What this means for Microsoft’s strategy

Build 2026 is more than a conference; it is a statement. By relocating to San Francisco and zeroing in on AI agents, trust, and a platform overhaul, Microsoft is telling the developer world that the next era of computing will be defined by autonomy, security, and a seamless blend of local and cloud AI. The company’s extensive investments in OpenAI, its own research, and the Azure infrastructure are converging into a single developer experience.

The shift also carries symbolic weight. San Francisco’s AI resurgence—boosted by startups and open-source communities—is a natural fit for a Microsoft that increasingly embraces open innovation. The same company that once battled the open-source movement now hosts the most popular developer platform on Earth (GitHub) and integrates with frameworks like LangChain and Semantic Kernel. Build 2026 will likely double down on that openness, welcoming contributions from across the AI ecosystem.

Looking ahead

As the industry grapples with the implications of AI agents that can reason, plan, and execute, Build 2026 stands to be the forum where Microsoft defines its role in that future. Developers who attend—whether in person or virtually—will leave with a clearer picture of how to build trustworthy, intelligent, and autonomous software on Microsoft’s expanding canvas.

The countdown to June 2, 2026, has begun. Expect more technical previews, community previews, and perhaps a few surprises that cement AI agents as the center of gravity for all things Microsoft.