The “open agentic web” was a major focus at the Microsoft Build conference, with dozens of new AI tools and features aimed at helping developers create autonomous systems that can shoulder routine and repetitive duties for human taskmasters and even undertake things like independent research with minimal supervision. The company unveiled a new GitHub Copilot coding agent that “supercharges” the AI assistant, allowing it to undertake more complex tasks and collaborate with other agents. Copilot Tuning was introduced as a “low-code” customization tool for adding corporate data to AI models. In all, Microsoft made more than 50 news announcements.

“We envision a world in which agents operate across individual, organizational, team and end-to-end business contexts,” Microsoft Chief Communications Officer Frank Shaw writes in the Microsoft blog, describing a future in which “the Internet is an open agentic web, where AI agents make decisions and perform tasks on behalf of users or organizations.”

VentureBeat reports the GitHub Copilot coding agent “can now operate as a member of software development teams, autonomously refactoring code, improving test coverage, fixing defects, and even implementing new features.” It can even collaborate with other agents, at all phases of the development cycle.

There are various plans available, including a free version with sign-up, while paid plans start at $10 per month or $100 per year.

Those who install GitHub Copilot in Microsoft’s Visual Studio Code will get GitHub Copilot Chat as an extension, “allowing the developer community to contribute to its evolution,” a move that “reflects Microsoft’s dual approach of both leading AI innovation while embracing open-source principles,” according to VentureBeat.

Microsoft’s Azure AI Foundry is a comprehensive platform designed to simplify the development, deployment, and management of AI applications and agents in a unified environment.

The Azure AI Foundry Agent Service is getting a general rollout, allowing developers “to build enterprise-grade AI agents with support for multi-agent workflows and open protocols like Agent2Agent (A2A) and Model Context Protocol (MCP),” VentureBeat writes, explaining that “this enables organizations to orchestrate multiple specialized agents to handle complex tasks.”

Copilot Studio, a low-code environment within the Azure AI Foundry ecosystem, has deployed Copilot Tuning to simplify using company data and workflows “to train models and create agents that perform domain-specific tasks with a high degree of accuracy using Copilot Studio — no coding required,” Microsoft explains in a blog post.

Multi-agent orchestration in Copilot Studio is now in public preview. The feature “enables agents to exchange data, collaborate on tasks, and divide their work based on each agent’s expertise,” Microsoft says, using the example of multiple agents collaborating across HR, IT and marketing to help onboard a new employee.

“Agent orchestration is becoming an increasingly important capability for enterprises as they need multiple agents to work in tandem to be able to achieve their desired autonomy in target processes and workflows,” writes InfoWorld.

Microsoft Corporate VP Copilot Studio Lili Cheng explores the topic in a blog post that includes highlights from Build.

A comprehensive Build 2025 recap — including keynote replays — is available at Microsoft News.

Related:
Microsoft Just Taught Its AI Agents to Talk to Each Other – and It Could Transform How We Work, VentureBeat, 5/19/25
Microsoft Debuts Windows AI Foundry for Local Model Development on AI PCs, SiliconANGLE, 5/19/25
NLWeb Is Microsoft’s New, Open-Source Tool That Integrates Generative AI Search into any Website, SiliconANGLE, 5/19/25