One Platform for Enterprise AI
Microsoft announced a significant restructuring of its cloud AI offerings, collapsing what had grown to more than a dozen distinct Azure AI service names into a single surface: Azure AI Foundry. The rebrand takes effect immediately, with the unified console rolling out to all Azure tenants over the next 30 days.
The consolidation addresses a consistent complaint from enterprise customers, who reported confusion navigating overlapping services — Azure OpenAI Service, Azure Machine Learning, Prompt Flow, Azure AI Studio, and several Cognitive Services products — each with separate billing meters, API endpoints, and compliance certifications.
What Changes for Customers
Existing API endpoints remain functional, so production workloads require no immediate migration. The change is primarily at the management and billing layer: a single Foundry dashboard now shows model deployments, token consumption, fine-tuning jobs, and evaluation pipelines under one pane.
New capabilities launching alongside the rebrand include a Model Catalog expansion to 1,800 open-weight models, a native prompt testing harness, and a compliance report generator that maps AI workloads to EU AI Act and NIST AI RMF frameworks automatically.
Competitive Positioning
The move mirrors Google's consolidation under Vertex AI and Amazon's push to centre enterprise AI on Bedrock. Microsoft's differentiator, according to the announcement, is the depth of Microsoft 365 and Dynamics integration — allowing Foundry-hosted models to read and write to organizational data without data leaving the Azure compliance boundary.
Pricing remains unchanged for the first 90 days, after which Microsoft will introduce a unified Foundry compute unit (FCU) that replaces per-service token meters.