The AI Wire

Lead dispatchazure ▸ development· filed 6 Sept 2026

Microsoft Foundry details six-phase process for migrating LLM models

Microsoft Foundry has detailed a six-phase process for moving LLM-powered applications between model generations, with model retirement dates now built into the platform lifecycle. The guidance emphasizes that swapping model names can silently change response formatting, schemas, and tool behavior, and that validation and canary rollouts are necessary because offline evaluation can miss production issues. A worked example ends in a deliberate split deployment rather than a clean cutover.

Read the dispatch →

Latest on the wire73 dispatches · page 01 / 08

72

simonwillison.net ▸ product· filed 31 Aug 2026

Simon Willison's hands-on test of ChatGPT Work: open internet access, browser automation, and unanswered safety questions

Simon Willison's hands-on test shows ChatGPT Work is an agent platform with unrestricted internet access, a headless Chrome browser, and parallel sub-agents, but OpenAI has not explained how it defends against prompt injection. The documentation is so sparse that he had to ask the agent itself to list its 223 tools and 44 skills. This leaves core safety mechanisms unaddressed in OpenAI's marketing.

71

youtube.com ▸ models· filed 30 Aug 2026

GLM 5.3 Flash: Z.ai’s cheaper, multimodal sibling matches larger models but slows on max reasoning

Z.ai's GLM 5.3 Flash is a 320-billion-parameter multimodal model that costs about one-ninth the output price of the flagship GLM 5.3. Independent tests show strong function-calling and agentic behavior, but max‑reasoning mode slows throughput to 50–60 tokens per second, and agentic fallbacks to the larger GLM are expensive, pushing users toward other frontier models for critical tasks.

70

youtube.com ▸ industry· filed 30 Aug 2026

Andrew Ng: Big AI labs are fear-mongering to shape regulation

Andrew Ng contends that leading AI companies are exaggerating risks to drive regulation that favors incumbents, while dismissing fears of a job apocalypse and warning that AI models hinder learning. He advises individuals to adopt AI skills and focus on product decisions rather than fear.

69

azure ▸ azure· filed 30 Aug 2026

VNet integration for Azure SRE Agent now generally available

Azure SRE Agent's VNet integration is now generally available, routing agent egress through customer VNets so it can reach private endpoints without public exposure. The feature inherits existing routing, NSG, firewall, and DNS policies, and offers an audit view for policy decisions, though that view is not a complete network audit trail. Setup requires a dedicated empty /27 subnet delegated to Microsoft.App/environments, and switching modes requires disconnecting the VNet first.

68

azure ▸ product· filed 30 Aug 2026

Grok 4.6 available in Microsoft Foundry public preview

SpaceXAI's Grok 4.6 is now available in Microsoft Foundry's public preview, aimed at long-horizon agentic work like repository issue resolution and business deliverables. Provider-reported benchmarks show strong performance on terminal and 3D modeling tasks, but production reliability remains unproven outside controlled evaluations.

67

azure ▸ product· filed 30 Aug 2026

Azure API Management Standard v2 Adds Availability Zone Redundancy

Microsoft has added Availability Zone redundancy to the Azure API Management Standard v2 tier, allowing customers to distribute gateway capacity across multiple zones without upgrading to a higher-priced tier. The platform handles failover automatically, but the protection covers only the gateway layer, so backend services and disaster recovery still require separate planning. The feature is enabled during service creation, with redeployment needed for existing instances.

66

azure ▸ product· filed 30 Aug 2026

Cohere Parse Document Extraction Model Now Available in Microsoft Foundry

Cohere Parse, a vision language model for extracting structured data from documents, is now available in Microsoft Foundry on Azure, converting tables, forms, and embedded images into Markdown while keeping inference within the customer's tenant. The model returns bounding boxes for downstream citation and reports throughput of 4.5 pages per second per GPU at $1.50 per thousand pages. Its domain-focused training on finance, insurance, scientific research, and nine languages means teams with other document types should validate output on their own corpus.

65

azure ▸ product· filed 30 Aug 2026

Azure DevOps plugin for GitHub Copilot adds pull request and work item management

Microsoft has released an Azure DevOps plugin for GitHub Copilot that lets developers view and manage assigned pull requests and work items without leaving the Copilot interface. The initial release is limited to opening, editing, and completing work items and pull requests from a "My Work" view, with Microsoft indicating plans to add more capabilities over time. Setup requires signing in and providing organization and project details, but the plugin currently lacks clarity on multi-organization support and real-time updates.

64

azure ▸ product· filed 30 Aug 2026

Microsoft Adds Cost Management Tools to Azure ARM MCP Server

Microsoft has added Cost Management tools to its Azure Resource Management MCP server, enabling AI agents to query cost data, price sheets, and budgets. The default tools cover narrow query and pricing scenarios, while an optional toolset requires opt-in via a header. Deployment involves some configuration friction, and Microsoft says it is expanding these capabilities.

63

azure ▸ product· filed 26 Aug 2026

Live Reports in Azure SRE Agent now in public preview

Microsoft has added Live Reports to its Azure SRE Agent, now in public preview, letting operations teams describe dashboards in natural language and receive a fixed-layout page that refreshes its underlying data on each open. The design separates authoring from rendering, so token costs apply only when reports are created or edited, and optional AI-powered interpretation or action buttons are per-report opt-ins. A four-layer security model restricts network access and validates tool calls against per-version allowlists.