For two years, the enterprise AI pitch has been about assistants: tools that draft, summarise and answer. At its annual Ignite technical conference on Wednesday, Microsoft made clear that the next phase is different. AI agents are now being treated as working colleagues, and they are expected to show results on the balance sheet.

Chief Executive Satya Nadella set out that thinking in a post on X. "Our ambition with everything we are announcing at Ignite today is to parlay technological advances into real outcomes and new points of leverage for people, organisations and society at large," he wrote.

The key word is "outcomes." Many companies are still trying to work out what their AI spending has actually returned, and Microsoft's announcements this year read like an answer to that question.

The Agent Becomes a Team Member

Copilot Workspaces is at the centre of the slate. Microsoft describes it as a platform where people and AI agents work together with shared context toward common objectives. That moves Copilot from a sidebar helper to something closer to a shared desk, where an agent can see the same project picture as the humans it works with.

Agent 365 Business Analytics may matter even more to the people who sign the cheques. Nadella called it an ERP system for AI, meaning a tool that helps companies tie AI spending to business outcomes and get better returns on those investments. Enterprise AI budgets are drawing more scrutiny from CFOs, so a product built to justify that spending is a deliberate move.

Microsoft also launched its own models: Microsoft-Decision 1.1, Code 2 and Cyber 2 Flash. The company says Cyber 2 Flash offers security capabilities that outperform frontier models at a lower cost. At the same time, it kept its platform open to other providers by bringing OpenAI's GPT-6.2 and Meta's latest models to Foundry. Microsoft is building its own models and hosting its rivals' at the same time.

On security, Nadella said Microsoft Perception is now generally available. It lets security teams and AI agents work together to identify and respond to threats at machine speed. Microsoft also announced the AI Transparency Initiative, which brings together more than 40 organisations to develop open standards for accountable, verifiable and safe AI systems.

Under the hood, HorizonDB, Microsoft's distributed PostgreSQL service on Azure, is now generally available. The company says it can deliver up to eight times the performance of self-managed PostgreSQL deployments. Nadella added that core Azure computing, storage and networking are being rearchitected to support AI agents operating at scale.

The Outage in the Background

The timing is awkward. Less than a week ago, a major Microsoft 365 disruption hit Teams, OneDrive, SharePoint and Office applications worldwide. Microsoft said the outage was caused by a combination of a hidden software flaw, a surge in traffic and reduced server capacity.

That matters for a company asking businesses to let autonomous agents into their core workflows. Promising machine-speed automation is easier when the basic services stay up. Microsoft listed customers and partners including Alaska Air, AT&T, BlackRock, BNY, Chevron, Epic, Publicis Groupe, Salesforce and Unilever. Those names suggest large enterprises are still on board, but how reliably the systems run will matter as much as the launches themselves.

Ignite 2026 shows that Microsoft no longer sees AI as an add-on feature. It wants agents built into the core of how businesses work, and to win that trust, they will have to prove both that they pay off and that they stay running.

With inputs from IANS