Copilot Expansion
The latest Copilot Expansion coverage — news, analysis, and updates from the WindowsNews.AI desk.
Anthropic Launches Claude Design Beta with WYSIWYG Canvas and Code Handoff for Paid Users
Anthropic has launched a beta of Claude Design, adding design system imports, a WYSIWYG visual canvas, and direct handoff to Claude Code for paid users. The update aims to streamline the design-to-code workflow directly within the Claude chat interface, with early adopters reporting significant time savings in frontend prototyping.
Barracuda Launches AI-Powered Post-Delivery Email Defense for Microsoft 365 and Google Workspace
Barracuda Networks launched Barracuda Integrated Email Protection in June 2026, a cloud-based, AI-powered email security service that focuses on post-delivery threat detection for Microsoft 365 and Google Workspace. The service uses advanced machine learning to scan emails after they reach inboxes, catching sophisticated phishing, BEC, and zero-day attacks that bypass traditional filters. It integrates via APIs for rapid deployment and offers automated remediation, user reporting, and continuous AI model updates to protect organizations against an evolving threat landscape.
Noam Shazeer's OpenAI Leap: How a Gemini Architect Jumped Ship in the AI Talent War
Noam Shazeer, a key architect of Google's Gemini AI and co-author of the Transformer paper, has reportedly left Google DeepMind for OpenAI in June 2026, less than two years after returning from his startup Character.AI. The move intensifies the talent war between the two AI giants and could impact the development of future models like Google's Gemini Nova and OpenAI's next-generation systems.
OpenAI's ChatGPT Tasks Beta Brings Scheduled AI Reminders to Windows — Here's How It Works
OpenAI launched the beta of ChatGPT Tasks on January 14, 2025, allowing paid subscribers to schedule reminders and recurring prompts. The feature turns ChatGPT into a background assistant, with deep Windows integration via the desktop app, push notifications, and up to 10 concurrent tasks. While still limited, Tasks signals OpenAI’s push toward proactive AI agents.
OpenAI Signs South Korea AI Safety Pact — Windows Admins Face New Compliance Reality
South Korea and OpenAI have signed a landmark AI safety agreement, joining the U.S., UK, and Japan in formal collaboration on model evaluation and threat intelligence. The deal has immediate implications for Windows IT admins managing Copilot, Azure OpenAI, and enterprise compliance, signaling a future where AI governance is as critical as traditional security policy.
Microsoft Pivots Copilot Cowork to Consumption-Based Pricing, Quietly Tests Cheaper DeepSeek Backend
Microsoft plans to shift Copilot Cowork to usage-based pricing starting June 2026, moving away from per-user subscriptions. The company is also reportedly testing a Microsoft-hosted, fine-tuned DeepSeek V4 model as a lower-cost optional backend for enterprise customers, aiming to offer more flexible cost structures and compete with open-source AI alternatives.
Microsoft Frames Trust as Infrastructure for Autonomous AI Agents in Asset Management
Microsoft's new whitepaper argues that trust must be engineered as a foundational layer for autonomous AI agents in asset management, integrating directly with Azure and Microsoft 365. The framework ensures every agent action is authenticated, logged, and policy-compliant, promising to accelerate enterprise adoption while mitigating risks.
Microsoft Drops Snapdragon X2 Surface Lineup, Claims Arm PC Parity with Intel
Microsoft launched a trio of Snapdragon X2-powered Surface devices—Surface Pro 13-inch, Surface Laptop 13.8-inch, and Surface Laptop 15-inch—on June 16, 2026, with immediate U.S. consumer availability. The new Arm chips bring native app support, local AI processing, and long battery life, positioning Windows on Arm as a genuine rival to Apple Silicon and Intel-based PCs.
Microsoft Debuts Copilot Cowork in General Availability, Shifting to Consumption Pricing and Model Flexibility
Microsoft has made Copilot Cowork generally available, introducing usage-based billing and model choice for its agentic AI platform. The move aims to give enterprises more control over AI costs and flexibility in selecting underlying models. New cost governance tools help manage spending, while the global rollout positions Microsoft at the forefront of autonomous workplace agents.
ChatGPT Scheduled Tasks Gets Dedicated Page After Reliability Overhaul
OpenAI rolled out a redesigned Scheduled Tasks experience for ChatGPT on June 17, 2026, featuring a dedicated page on web and mobile for managing automated prompts and reminders. The update addresses past reliability issues with improved backend infrastructure, granular notifications, and webhook support, offering Windows IT pros a powerful new automation tool.
Microsoft Copilot Cowork Goes GA with Consumption-Based Pricing and Granular Agent Governance
Microsoft released Copilot Cowork to general availability in mid-June 2026, transitioning the agentic AI tool from preview to a paid service with consumption-based pricing and extensive governance controls. The launch gives enterprises the ability to automate complex, multi-step workflows across Microsoft 365 while maintaining security, compliance, and cost management through new policy, audit, and DLP integrations.
Microsoft Turns Teams Rooms Into AI Workflow Hubs at InfoComm 2026
Microsoft showcased at InfoComm 2026 how Teams Rooms, powered by Copilot Studio and AI voice agents, can automate meeting workflows and serve as an intelligent layer for office productivity. The company also extended AI to Teams Phone and revealed new management tools for pro AV integrators, signaling a shift from passive communication rooms to active workflow platforms.
Microsoft’s Azure AI ‘Middle Layer’ in China: How It Gives ByteDance and Tencent Access to OpenAI Models
A Bloomberg investigation reveals Microsoft has built a ‘middle layer’ within Azure China that gives ByteDance, Tencent, and other giants access to OpenAI models while skirting U.S. export controls and Chinese data laws. The secretive business, worth hundreds of millions, raises serious geopolitical and compliance questions about the flow of advanced AI into China.