streugut.ch
Archive

The direction

Where the AI race is heading

AI competition is moving away from a single frontier race and toward control of the path from model to deployed workflow. That shift is being carried by enterprise rollouts through implementation partners, by agents embedded inside governed software environments, and by a compute contest widening from chips to cloud capacity. What is particular now is how distribution, deployment, infrastructure, and compliance are locking together into the practical machinery of scale.

The move

What actors actually did

Microsoft makes Copilot agents available for Cloud PCs and launches new SRE agent tools

Microsoft brought several new developments in AI automation this week. On May 27, it released Windows 365 for Agents in public preview, which lets Copilot agents run workflows in Cloud PCs, handling both legacy and UI-based systems under enterprise controls. Building on this, Microsoft launched SRE Agent tools in Azure MCP Server, allowing developers to manage Azure SRE Agents from any IDE, terminal, or AI assistant, providing more flexibility for managing cloud resources.

Elsewhere, Alibaba Cloud announced a broad agentic AI ecosystem on May 26. The update included new products like the Qwen3.7-Max language model, high-performance chips, upgraded infrastructure, and an AI-native platform for global customers. Meanwhile, OpenAI and Thrive Holdings rolled out Tax AI for Crete accounting firms on May 27, automating tax return preparation for 7,000 returns and reporting higher accuracy rates.

Snowflake also took steps in the cloud and enterprise AI field. On May 27, they signed a multi-year collaboration with AWS, putting $6 billion toward speeding up enterprise adoption of agentic AI. They further announced the intent to acquire Natoma to bring secure AI connectivity from Natoma’s MCP platform to their customers.

Sources