Browse All Artificial Intelligence Content (832)

Stefan Kingham’s documentary traces how Visual Studio Code evolved from early web-based editor experiments into a widely used, open-source developer tool, covering key technical and product shifts like Electron, extensions, LSP, remote development, and the editor’s adaptation to the AI era including GitHub Copilot.

GitHub explains the “AI harness” concept for guiding LLM execution, and how harness engineering adds the runtime, tools, and guardrails that let AI agents work safely against a codebase in agentic workflows.

sbaynes introduces MAI-Transcribe-2, a speech-to-text model positioned for low-latency, multilingual transcription with features like diarization, word-level timestamps, keyword biasing, and configurable output styles, alongside benchmark results and launch pricing aimed at high-throughput transcription workloads.

Kayla Cinnamon explains how the GitHub Copilot app can run multiple agent sessions in parallel on the same project by isolating each session with its own Git worktree, so you can queue up tasks (like feature work, accessibility review, and tests) and review results as they complete.

Microsoft Security Research (Noam Kochavi and Sarah Wolstencroft) explains how “ASCII smuggling” with invisible Unicode tag characters moved from AI prompt-injection research into real-world phishing evasion, and shares practical detection and mitigation guidance for email defense pipelines.

Allison announces that GitHub is gradually reopening sign-ups for Copilot Business and Copilot Enterprise paid by credit card or PayPal, alongside billing and account-vetting changes that affect how seats are activated, charged, and how overage usage may require additional payment.

dotnet covers the production-readiness pieces for an AI agent, focusing on observability with OpenTelemetry, governance and data protection with Microsoft Purview, and quality evaluation as the SDK evolves, then shows how to deploy the agent to Foundry Hosted Agents and maintain it with a refresh-prompt workflow.

dotnet demonstrates how to make an AI agent interact with real user data more safely by adding file access, explicit human approvals for sensitive actions, and durable memory that persists across restarts.

dotnet explains four additions that make an AI agent more practically useful: skills for loading domain knowledge on demand (including centrally managed skills in Azure AI Foundry), constrained shell access for file work, a CodeAct sandbox for writing and running code, and background agents for parallel research.

dotnet demonstrates how to stand up an agent “harness” in C# using a single AsHarnessAgent call, then incrementally add tool calling, web search, and planning so a vague request becomes a tracked, step-by-step execution loop.

Microsoft Developer demonstrates building the “Sparkles” AI agent in VS Code using the Microsoft Foundry Toolkit, including deploying a model, wiring up an MCP server, and validating the agent with traces and evaluation runs for repeatable testing.

Cassidy Williams breaks down the new AI vocabulary showing up in developer conversations—loops, harnesses, squads, hill climbing, and model “openness”—and explains how these concepts map to practical ways teams build and improve agent-driven workflows (with GitHub Copilot as a concrete example of a “harness”).

Allison announces an update to GitHub Copilot enterprise-managed settings that lets organizations set any Copilot model as the default for new conversations, with optional per-team overrides via managed settings files.

Allison announces general availability of content exclusion support in the GitHub Copilot app and Copilot CLI, so enterprise, organization, and repository admins can prevent Copilot from using specified files as context and better protect sensitive code during agentic workflows.

Erik Kristensen explains how GitHub Copilot reduced end-to-end cost for agentic coding by optimizing whole tasks (not individual tool calls), including selective compression of noisy command output, removing wasted formatting, tightening prompts with regression tests, and batching background results to avoid extra model turns.

John Maeda shares a practical take on what’s genuinely new in AI: setting a clear goal, defining how you measure progress toward it, and using iterative loops that “hill climb” toward better results as tokens get cheaper and more capable.

GitHub’s podcast episode with Cassidy, Marlene, and GPS breaks down emerging AI terms—loop engineering, harness engineering, and squads—and connects them to practical agentic workflows, including how iteration loops affect cost and performance and how “open weight” differs from truly open source AI.

The Visual Studio Code Team shares Insiders updates for VS Code 1.137, focusing on Copilot-related workflow improvements such as clearer model token limits, better context handling when pasting GitHub issue/PR links, and updated new-session composer controls.

James Montemagno introduces VS Code agent plugins and how they package and share AI customizations—like skills, MCP servers, and custom agents—so teams can reuse the same setup across projects and environments.

AmirJafari announces general availability of the Microsoft Fabric data agent integration in Copilot Studio, enabling agents to call governed Fabric data as a tool (Fabric IQ Data MCP) and return permission-respecting, data-grounded answers. The post highlights the new tool-based authoring experience, improved orchestration, and expanded publishing targets.

GitHub hosts a multi-hour Copilot Day live session covering new releases, hands-on demos, and practical workflows, including how Copilot works across GitHub, the CLI, and VS Code, plus a live-coding segment with Burke Holland, Pierce Boggan, and guests.

John Savill gives a quick, practical explanation of why generative AI models sometimes “hallucinate” (confidently produce incorrect information) and why this behavior is a natural outcome of how models generate text from probabilities rather than retrieving verified facts.

Amit Ganguli announces the public preview of the Resiliency Agent in Azure Copilot, which helps teams assess zonal resiliency across subscriptions, prioritize fixes with cost-aware indicators, and generate deployment-ready Bicep/Terraform templates and scripts to remediate gaps.

Microsoft Developer demonstrates how to turn a deployed model into a reusable AI agent using the Microsoft Foundry SDK in Python, including adding instructions, running a multi-turn conversation, and wiring in MCP tools so the agent can take actions while maintaining context.

Allison announces a generally available GitHub billing update that lets admins set an expiration date on individual user budgets, so temporary budget overrides are removed automatically and users fall back to the next applicable budget.

Allison announces a public preview update to GitHub Copilot code review that adds approval assessments on pull requests and, when enabled by admins, allows Copilot to submit an approval that can count toward required-approvals rules.

Microsoft Developer presents a student-focused course that uses the GitHub Copilot app to turn a resume into a live portfolio site, while teaching Git fundamentals, GitHub Pages deployment, Copilot modes, context/tokens, MCP servers, skills, and bring-your-own-model with Azure AI Foundry.

GitHub runs a beginner-friendly virtual training on the GitHub Copilot SDK, focusing on how developers can programmatically interact with Copilot and build custom agent capabilities that fit their team’s workflows across multiple languages.

Jonah Andersson explains how to take AI agents from prototype to production by building MCP servers on Azure Functions, using .NET to expose tools that agents can discover and invoke at runtime. The session uses a DevOps CI/CD scenario where agents detect deployment issues, diagnose failures, and create rollback PRs in Azure DevOps.

Aaron Powell walks through reviewing GitHub pull requests directly inside Visual Studio, from opening the PR list and navigating diffs to leaving inline comments, approving, and merging. He also shows how to bring GitHub Copilot into the review by adding a PR as chat context using the UI or the #pullrequest: command.

Virginia_R announces a Preview feature in Fabric data agents that upgrades natural-language-to-DAX translation for Power BI semantic models using an agentic, iterative approach. The post explains how planning, tool use, intermediate-result inspection, and instance value indexing improve filter accuracy, consistency across Microsoft experiences, and overall query reliability.

GitHub introduces canvases in the GitHub Copilot app and shows how beginners can use the /create-canvas skill to generate custom, interactive workflow surfaces like release checklists, kanban boards, and triage views, then iterate on them and share state in real time.

Allison announces that Anthropic’s Claude Fable 5.1 is now available as a selectable model in GitHub Copilot, including details on where it’s available, how admins enable it for Business/Enterprise, how it’s billed under usage-based pricing, and the default data-retention requirement plus ZDR options for eligible enterprises.

John Edward covers a Copilot Studio authentication update that adds Microsoft Entra ID-based identity for both delegated user access and service-to-service (S2S) scenarios, aimed at making enterprise agent deployments easier to secure and integrate with existing applications.

Allison explains a GitHub Copilot change for GitHub Team plans: when a user has Copilot seats in multiple organizations, model access is now determined only by the organization that is billed for the user’s Copilot usage, aligning governance and billing.

Microsoft Developer shares a quick developer news roundup covering Bring Your Own Model entering preview in Visual Studio, updates to the model router in Microsoft Foundry Models (including region expansion and refreshed model pool), and the announced date for .NET Conf 2026.

Allison announces that several AI models are deprecated across GitHub Copilot experiences as of September 1, 2026, and outlines the replacement models plus the admin steps Copilot Enterprise customers may need to take to enable alternatives via model policies.

FabSubhojit announces the general availability of Fabric Migration Assistant for SQL database, a guided workflow for assessing schema compatibility, provisioning SQL database in Fabric, deploying schema with fix suggestions, and copying data. The update adds a Validate step that analyzes a DACPAC before any Fabric resources are created.

Lily Ma explains how to extend Azure SRE Agent with operational tools by hosting Model Context Protocol (MCP) servers on Azure Connector Namespace. The post walks through deploying an Azure SQL MCP server with azd, wiring it into SRE Agent as an MCP connector, and securing access using managed identity.

Filisha Shah explains what to focus on as more application logic shifts into AI agents, highlighting why observability and evaluation become core skills and how Azure AI Foundry brings datasets, evaluations, and monitoring together for keeping agentic systems healthy.