Browse Artificial Intelligence News (436)

Allison summarizes weekly GitHub Copilot updates across the Copilot app, Copilot CLI, and VS Code, focusing on session workflows, side questions, model/credit visibility, and new review and feedback capabilities that help developers stay in context while iterating on changes.

Allison announces a new “Potential return on investment” section in the GitHub Copilot impact dashboard, helping admins relate Copilot spend (based on AI credit consumption) to pull request output and adoption phases, with updated cohort counting across the full 28-day reporting window.

Allison announces general availability of Lite and Balanced effort levels for GitHub Copilot code review, letting teams tune review depth per pull request and set organization-wide defaults that repositories can inherit.

Allison announces an update to the GitHub Copilot usage metrics API that adds per-agent reporting for third-party agent apps (such as Claude and Codex) used in GitHub workflows. The change introduces a new totals_by_3rd_party_agent array and clarifies how to interpret the new counters alongside existing telemetry fields.

Allison explains a GitHub Code Quality change: enabling Code Quality no longer creates a ruleset that automatically requests GitHub Copilot code review on pull requests, and existing auto-created rulesets have had those settings turned off.

Allison announces a generally available GitHub Copilot enterprise control that lets owners allowlist or denylist Model Context Protocol (MCP) servers via managed settings, helping teams approve trusted MCP endpoints and block untrusted or non-compliant ones across supported Copilot clients.

sbaynes outlines an Azure-native reference architecture for deploying Fireworks AI open-model inference on Azure AI Foundry, with guidance on moving from prototype to production while tracking latency, token usage, and cost. The post highlights optional platform components for caching, rate limiting, secrets management, and observability.

Jacklyn Carroll explains how slash commands in the GitHub Copilot app speed up common workflows like planning, pressure-testing designs, and delegating multi-step implementation work, with concrete examples you can paste into Copilot to get started.

Walker Chabbott announces that the open-weight Kimi K3 model is now generally available in GitHub Copilot, including where it can be selected, how it’s billed under usage-based pricing, and what Copilot Business/Enterprise admins need to do to enable it.

Mike Hulme explains why Microsoft was named a Leader in Gartner’s 2026 Magic Quadrant for AI-augmented code modernization tools, and outlines how GitHub Copilot modernization and Azure are positioned to assess, upgrade, and migrate legacy applications with agentic workflows while keeping developers in control.

Ran Rosin summarizes why KuppingerCole named Microsoft a Leader in its 2026 CNAPP Leadership Compass, and what that says about where cloud security is heading: unified risk management across cloud infrastructure, identities, data, applications, and AI, with runtime-driven prioritization and SOC integration.

Dan Hellem announces the general availability of the Azure DevOps Remote MCP Server, explaining how AI assistants can securely connect to Azure DevOps via a hosted endpoint. The post covers the minimal mcp.json setup, Entra ID authentication requirements, current client support, and when to use the local MCP server instead.

Bruno Capuano shows how to build a Windows-only C# console app that performs live speech-to-text locally using Foundry Local and a compact Nemotron streaming ASR model, including model lifecycle management (download/cache/load/unload) and real-time interim/final transcription via async streams.

Allison announces that the GitHub Copilot Billing Preview app has been retired and explains how to manage Copilot spend directly in GitHub billing settings, including viewing AI usage, setting budgets, and exporting or pulling raw usage data via reports and the billing API.

Bria Ford shares how GitHub’s legal team used GitHub Copilot CLI to turn repetitive legal work into repeatable, repo-backed workflows—ranging from contract drafting style guides to DMCA code-analysis playbooks—while keeping human judgment central.

Ron Pessner announces updates to Microsoft’s Zero Trust for AI approach, including new AI-focused checks in the Zero Trust Assessment, a DevSecOps pillar in the Zero Trust Workshop, and a companion e-book with practical patterns for securing AI agents, AI memory, and the software supply chain.

Giles Odigwe announces the stable release of the GitHub Copilot Agent for Microsoft Agent Framework in .NET and Python, showing how to use Copilot’s coding harness (shell, file, URL, and MCP tools) with Agent Framework features like middleware, streaming, OpenTelemetry tracing, and human-in-the-loop approvals.

Julia Muiruri explains how to turn a single, oversized AI-generated pull request into a clean stack of smaller PRs using GitHub stacked pull requests and the gh-stack CLI, with a practical example and reviewer-focused guidance for keeping CI and merge rules working across the whole stack.

Allison announces the retirement timeline for GitHub Spark on github.com, including key dates, what happens to existing deployed apps, and what builders need to do to keep editing their apps after shutdown. The post also explains the impact of the earlier GitHub Models retirement on Spark apps that call llm().

sbaynes reports on PRISM2, a pathology foundation model from Microsoft Research and Paige (Tempus) that links tissue images with diagnostic language from real pathology reports, aiming to reduce the need to rebuild separate models for each cancer-detection task.

Allison announces a public preview feature in GitHub Code Quality that can generate a code coverage workflow with AI from repository settings, opening a pull request for review before you merge.

Allison announces that GitHub Copilot cloud agent now lets you choose a reasoning level (for supported models) when starting a task, trading off deeper reasoning against higher token and credit consumption across paid Copilot plans.

Allison announces a GitHub Copilot Enterprise governance update that lets enterprise admins target managed settings to specific enterprise teams using itemized configuration files, so large orgs can scale guardrails while still letting teams tailor Copilot within defined boundaries.

Ricardo M. Martins describes a reference Azure-based operations assistant that uses RAG over runbooks and gates any write action (like incident creation) behind explicit user confirmation, with practical notes on Azure AI Search ingestion pitfalls, identity boundaries, managed identity, and idempotency.

Allison announces a new GitHub Copilot capability: Copilot cloud agent automations can now be triggered by issue comments or pull request comments, enabling comment-driven workflows like generating documentation, investigating errors, and creating follow-up tasks.

Allison announces upcoming GitHub Copilot model deprecations effective September 1, 2026, including the impacted Copilot experiences and the recommended replacement models. The post also calls out an exception for some annual individual subscribers and explains what Copilot Enterprise admins may need to change in model policies.

Microsoft Threat Intelligence details CaptiveCrunch, a Storm-2945 (Midnight Blizzard) campaign abusing captive portals to manipulate traffic, deliver malware, and steal credentials and tokens. The post breaks down the CornFlake and ChocoShell toolchain, device code phishing against Microsoft Entra ID, and provides Defender/Sentinel hunting queries plus practical mitigations for travelers and enterprises.

Allison announces the deprecation of the Gemini 2.5 Pro and Gemini 3 Flash models across GitHub Copilot experiences, and explains what Copilot Enterprise admins may need to change in model policies so users can select supported alternatives in VS Code and on github.com.

Allison announces a public preview for GitHub Enterprise that lets Copilot admins target AI model policies to enterprise teams, so different groups can be granted additional models beyond an enterprise-wide baseline.

Samantha Kubota introduces Echoverse, a set of high-fidelity synthetic “worlds” for training computer-use agents, and explains why depth (stateful workflows and database-grounded grading) matters more than sheer environment count. The post shares results from supervised fine-tuning and reinforcement learning, plus links to released code and datasets.

Amaury Levé introduces an open-source polyglot unit-test generation agent that learns a repo’s conventions, writes tests, and validates they build and run in the project’s normal test workflow. The post explains the agent’s research/plan/execute/check loop, shows benchmark results versus stock Copilot, and includes setup commands for Copilot CLI and VS Code.

Kristen Womack’s July 2026 roundup for Azure Developer CLI (azd) summarizes releases 1.27.0 through 1.29.0, including new extension install options, a full tool uninstall workflow, Azure AI Foundry resource modeling in azure.yaml, container deployments to Azure App Service, and CI-friendly non-interactive behavior plus a breaking flag rename.

diptiborkar summarizes why Microsoft Fabric and OneLake were recognized as a Leader in Forrester’s 2026 Data Lakehouses report, focusing on a unified lakehouse foundation for analytics and AI, open table formats, cross-cloud interoperability, and built-in governance and security across engines like Spark, SQL, KQL, and Power BI.

Allison announces the retirement of GitHub Models as of July 30, 2026, including the shutdown of the playground, model catalog, inference API, and BYOK support. The post also points teams to Microsoft Foundry for model access and GitHub Copilot for building AI-powered workflows on GitHub.

Samantha Kubota explains how Microsoft is pushing better AI performance per token and per dollar by co-optimizing models with evaluation harnesses and RL environments, sharing recent “Flash” model releases and benchmark results that reduce GPU costs while maintaining quality across products.

Cassidy Williams shows how she used the GitHub Copilot app to modernize an old React codebase by splitting work into stacked sessions and stacked pull requests, keeping changes reviewable while iterating on plans, branching strategy, and dependency upgrades.

Shawn Henry and Tamir Dresher show how to run a multi-agent “Squad” team on top of the GitHub Copilot CLI/SDK and expose it as a first-class Microsoft Agent Framework (MAF) AIAgent in a .NET app, including DI wiring, session configuration, and OpenTelemetry-based observability.

Alym Rayani rounds up July 2026 updates across Microsoft Security, focusing on securing AI and agent environments. The post highlights Project Perception’s agentic SecOps approach, new Microsoft Defender protections (including prompt-injection defenses and cloud agent runtime coverage), Entra identity updates like passkeys-by-default, and Purview controls for data protection and investigations.

Allison summarizes the July 2026 updates for GitHub Copilot in Visual Studio, including a new preview agent built on the Copilot SDK, built-in .NET and Azure skills, selected-code review actions, and organization-wide custom instructions for consistent Copilot behavior across teams.

Allison announces a new GitHub Copilot enterprise managed setting that lets organizations restrict which managed devices can host remotely controlled Copilot sessions, including options to require SSO, disable remote control, or allow it without restrictions.