Microsoft

Data Formulator 0.7 is an open-source AI-powered system for enterprise data analytics that combines data connectivity, agent-guided exploration, and visualization refinement in a shared workspace.

Microsoft's MAI-Image-2.5 ranks third on Arena's text-to-image leaderboard, on par with Google's Nano Banana 2 but still behind OpenAI's Image-2. The model shows clear gains over its predecessor, especially in rendering text inside images and commercial visuals.

Modern AI systems are powerful not because they replicate human intelligence, but because they extend structures already present in human cognition and language. This perspective explains AI's capabilities and limitations, and reframes AI safety as a system-level challenge requiring engineering and governance, not fear of rogue AI.

theta is a Rust CLI that manages agent configurations by reading a theta.toml file, resolving, locking, materializing, and casting them to any supported harness (e.g., Claude Code, Codex CLI, GitHub Copilot, Cursor). It works like a package manager for agent harness resources. Installation is straightforward, and it supports adding rules, tools, skills, and subagents, with validation and casting commands. The project is heavily inspired by uv and is the canonical implementation of the theta-spec.

A vulnerability in Microsoft Copilot Cowork allows attackers to exfiltrate OneDrive files through prompt injection and external images in automatically sent emails.

This PDF report from Microsoft Research analyzes global AI diffusion trends for Q1 2026, offering key insights and data. The full content is available in the original PDF document.

Over the weekend: Musk, Zuckerberg, and Sacks killed Trump's draft AI safety executive order in three Wednesday-night phone calls. Anthropic closed a $30B+ round the same Saturday — while Microsoft quietly cancelled its internal Claude Code pilot after token billing ate the entire annual AI budget, redirecting developers to Copilot. CISA logged 15,000 attacks on a same-week Drupal SQL flaw. The first cross-registry supply chain attack — TrapDoor — hit npm, PyPI, and Crates.io at once, using .cursorrules and CLAUDE.md config files as the carrier. And the White House personally overrode the Pentagon to keep Claude inside the NSA.

Microsoft is a key AI player with its OpenAI investment and growing cloud AI business, which achieved an annual revenue run rate of over $37 billion. Despite a recent 12% decline, the stock is a strong long-term buy due to deep integration with corporate customers and AI integration opportunities. At 25x forward earnings, it offers an attractive entry point.

When analyzing data, Microsoft Copilot invents country differences where none exist. Mathematician Adam Kucharski fed the tool identical datasets with different country labels, and Copilot delivered detailed stereotypes instead of accurate results. Thinking models catch the trick, but only if users know when to reach for them.

Microsoft Research introduces Webwright, a terminal-native browser agent framework that replaces click-trace web automation with reusable Playwright scripts. Using a single agent loop across three modules and roughly 1,000 lines of code, Webwright powered by GPT-5.4 reaches 60.1% on the long-horizon Odysseys benchmark and 86.7% on Online-Mind2Web — the highest AutoEval score among open-sourced harness recipes.

The Getac G140 is a rugged Windows 11 Pro tablet designed for harsh environments like fire/rescue, automotive, and utility work. It features an AMD Ryzen AI processor, hot-swappable batteries, and extensive port options. While extremely durable and capable of running Microsoft Copilot+ AI tools, it underperforms in benchmarks, is heavy at 3.95 pounds, and has a dim screen at 1000 nits. The device is ideal for niche industrial use but expensive (up to $4,000) and bulky.

Microsoft is canceling most Claude Code licenses due to high costs, shifting engineers to GitHub Copilot. Uber burned through its 2026 AI budget in four months. Experts note that while token prices drop, consumption surges, making AI more expensive than human labor in some cases.

GitHub Copilot now serves 140,000 organizations, with over 100% year-over-year growth. Gartner positions GitHub as a Leader for the third consecutive year, highest in ability to execute.

Microsoft Research released Fara1.5, a family of browser computer-use agents in 4B, 9B, and 27B sizes. Fara1.5-27B scores 72% on Online-Mind2Web, outperforming OpenAI Operator and Gemini 2.5 Computer Use. The release also includes FaraGen1.5, a synthetic data pipeline that trains agents on gated domains.

CopilotKit's 2026 releases include three tools: AG-UI protocol, AIMock testing suite, and Pathfinder knowledge server, addressing interaction, testing, and knowledge retrieval gaps for production-grade agentic AI. Adopted by major cloud providers and Fortune 500 companies.

The longtime partners expanded their relationship to accelerate enterprise AI adoption.

Anthropic's Mythos model has discovered thousands of severe security vulnerabilities, including many zero-day flaws that have gone undetected for decades. Banks worldwide fear cybercriminals will exploit this AI to rob them. Anthropic has granted access to a defensive coalition including Microsoft, but not to banks in Australia, the UK, or Europe.

Microsoft Research releases MagenticLite, an agentic application designed for small models, along with MagenticBrain orchestrator and Fara1.5 computer-use model. The system works across browser and local file system, achieving state-of-the-art results on web navigation tasks while keeping data on-device.

Vega is a new zero-knowledge proof system from Microsoft Research that enables users to prove facts from government-issued credentials without revealing the credential itself. It achieves under 100ms proving time on commodity devices using folding schemes, and is designed for real-world digital identity formats like mobile driver's licenses and the EU Digital Identity Wallet.

In 2026, enterprise agentic AI has moved from pilots to production. This guide ranks the top 10 platforms — Salesforce Agentforce, Microsoft Copilot Studio, ServiceNow, LangGraph, and more — with verified pricing, real adoption data, and honest constraints to help enterprise teams make the right platform decision.

Tamarillo analyzed ~400K public GitHub repositories containing configuration files for AI coding assistants (harnesses) like Cursor, Copilot, and Claude. The study covers market share, adoption dynamics, configuration surface anatomy, multi-harness co-occurrence, and repo demographics by stars, language, and owner type. It reflects configuration intentions and is a lower bound on actual usage.

Microsoft announced at Open Source Summit North America 2026 the public preview of Azure Linux 4.0 and general availability of Azure Container Linux, alongside contributions to the Agentic AI Foundation (AAIF) and open governance tools for agentic systems.

Remote control for GitHub Copilot CLI sessions is now generally available on github.com and GitHub Mobile. Developers can start a session in VS Code or the CLI, then monitor and adjust it from another device. Features include real-time monitoring, mid-flight instruction changes, permission approvals, and a seamless cross-device workflow, with privacy by default.

MLflow AI Gateway automatically logs LLM call traces, helping developers debug agentic apps and coding assistants. The article covers usage, integration with LiteLLM, support for Copilot CLI, and discusses tracing as a scaling problem.

Microsoft is abruptly canceling licenses for Anthropic's Claude Code in favor of its own Copilot CLI, revealing strategic prioritization of internal tools over partner products despite close ties with OpenAI. The shift may cause developer friction and highlights the tension between best-of-breed tools and corporate control over AI infrastructure.

Microsoft's open-source AI Engineer Coach analyzes AI coding assistant logs locally, providing practice scores, anti-pattern detection, and output metrics to help developers improve coding efficiency.

GitHub launches a standalone desktop app for Copilot, designed to manage coding agents, issues, pull requests, and development sessions from a single interface, directly competing with Anthropic's Claude Code and OpenAI's Codex. Built on Copilot CLI, the app offers a unified inbox, side-by-side diff reviews, session history, and multi-agent support. Currently in public preview.

LightningTrack is an issue tracker designed for AI-assisted development. It turns any issue into a context-rich prompt for Copilot, Cursor, or Claude with one click, supports email-to-issue conversion, sprints, custom fields, reports, and an agent-friendly API.

Microsoft Research clarifies the scope of its paper on AI delegation, noting that while models show fidelity degradation in long-horizon tasks, production systems mitigate these effects, and the benchmark is a diagnostic tool for future improvement.

A single ~7,600-line Markdown bootstrap file (AI_PROJECT_SETUP.md) automates the setup of AI coding assistants including Claude Code, ChatGPT Codex CLI, Cursor, and GitHub Copilot. It generates 13-section rules, safety hooks, session resume, dual-write memory, bilingual GitHub files, and more. Just download the file, tell your AI to read and execute it, and within 1-3 minutes you have a complete, gitignored AI tooling configuration.

A relatively quiet day in AI news highlights a smaller trend: the convergence of coding agent form factors around Conductor's pioneering approach. Key stories include GitHub's new Copilot App mimicking Conductor, OpenAI's Codex mobile launch, LangChain's agent infrastructure updates (SmithDB, Engine, Labs), Anthropic's Claude Code restrictions backlash, Figure's 24/7 autonomous sorting livestream, and notable research releases on diffusion LMs, time-series forecasting, and mechanistic interpretability.

JDS is a GitHub Copilot CLI plugin that enforces structured workflows (design before code, tests before implementation, evidence-based verification), transforming AI coding assistants from autocomplete tools into disciplined software engineers. It features a multi-stage pipeline (think → plan → execute → verify → finish) with both flexible and rigid skills for different task types.

Microsoft is planning to remove most Claude Code licenses by June 30, pushing developers to use GitHub Copilot CLI instead. The decision is driven by financial and strategic reasons, despite Claude Code's popularity. Anthropic models remain accessible via Copilot CLI.

With Edge's mobile app, you can now request summaries of multiple tabs, turn pages into podcasts, and tap into your browsing history and past chats.

Conductor is an open-source CLI from Microsoft that defines multi-agent workflows in YAML with deterministic routing instead of LLM-based dynamic orchestration, reducing cost and latency. It supports mixed models, parallel execution, human gates, script steps, and a web dashboard, ideal for structured workflows like code review and research synthesis.

Microsoft built MDASH, a system that pits more than 100 specialized AI agents against each other to find software vulnerabilities. On Patch Tuesday alone, the system uncovered 16 security flaws in Windows, four of them critical. Microsoft isn't saying which AI models power the system.

AI agents are vulnerable to 'whimsical' adversarial strategies that appear absurd to humans but reliably succeed. Microsoft researchers generated 30K such strategies from diverse Wikipedia articles, demonstrating that even frontier models like GPT-5 can be manipulated in negotiation environments. These out-of-distribution attacks exploit blind spots in safety training that focuses on human-perceptible threats.

Microsoft is upgrading Edge's Copilot AI chatbot so it can read all open tabs at once, compare products, and summarize articles. New additions include long-term memory, a tool that turns tabs into AI podcasts, and a quiz mode.

Microsoft Edge is adding a new feature that lets Copilot AI gather information from all open tabs. You can ask questions, compare products, and summarize articles. Microsoft is retiring Copilot Mode and folding its agentic capabilities into 'Browse with Copilot'. Other updates include a 'Study and Learn' mode, tabs-to-podcasts, AI writing assistant, browsing history access, long-term memory, a redesigned new tab page with Journeys, and mobile screen sharing.

Everything Claude Code is a comprehensive performance optimization system for AI agent harnesses, originally an Anthropic hackathon winner. It provides a complete system including skills, instincts, memory optimization, continuous learning, security scanning, and research-first development. The system supports multiple harnesses such as Claude Code, Codex, Cursor, OpenCode, Gemini, and GitHub Copilot, evolving over 10+ months of intensive daily use building real products. The latest v2.0.0-rc.1 introduces a dashboard GUI, operator workflows, and the ECC 2.0 alpha.

Microsoft releases a lightweight foundation model that can predict AC optimal power flow in milliseconds, boosting efficiency and unlocking cost savings in grid analysis.

Microsoft announced its new multi-model agentic security system (MDASH) has achieved top performance in an industry benchmark. The system, which coordinates over 100 specialized AI agents across frontier and distilled models, discovered 16 new vulnerabilities in the Windows networking and authentication stack, including four critical remote code execution flaws.

In the Musk v. Altman trial, Microsoft appears reluctant and detached. Their opening statement was essentially an ad for Microsoft products, implying the trial's absurdity. Despite being an early major investor in OpenAI, Microsoft was notably absent from key decisions. CEO Nadella called the OpenAI board drama 'amateur city,' and Microsoft lawyers repeatedly showed the company had no involvement in controversial events.

Researchers discovered a prompt injection vulnerability in VSCode Copilot's agent mode, allowing attackers to bypass user confirmation by exploiting a TOCTOU flaw in the applyPatch tool, leading to arbitrary file write and remote code execution via overwriting .git/config or shell configuration files.

In light of recent events, consumers are urged to wield their power against AI giants by replacing ChatGPT with Claude and avoiding Microsoft Copilot.

SubVault is an MCP server that gives AI tools like Claude, Cursor, and Copilot persistent memory by extracting structured knowledge from conversations. It remembers decisions, facts, people, and project context across sessions, scoring information by authority, recency, and relevance. Free during early access, setup takes 30 seconds.

RoBrain is an open-source shared memory for teams using AI agents. It automatically captures every decision and the alternatives ruled out, across all developers' sessions, and flags contradictions. Teams revisit past decisions with original rationale instead of re-litigating from scratch. Works with Claude Code, Cursor, and Copilot.

Microsoft Israel's top executive Alon Haimovich stepped down amid an internal investigation into the subsidiary's work with Israel's defense ministry. Reports indicate Azure cloud services were used for mass surveillance and AI-powered target selection in Gaza, leading to legal and transparency concerns.

Learn how one Hubber used GitHub Copilot CLI to build an extension that turns any codebase into a unique, roguelike dungeon.

Cplt is a sandbox wrapper for AI coding agents that provides kernel-level filesystem and environment isolation on macOS and Linux. It protects secrets by restricting access to credentials, SSH keys, cloud configs, and blocking lifecycle scripts. Supports Copilot, OpenCode, and other agents with fine-grained control over files, network, and environment variables.