# Microsoft Is Killing Copilot Features — Here’s Why > Source: > Published: 2026-08-13 19:08:57+00:00 Microsoft is shutting down Deep Research, AI-generated Podcasts, Group Chats, Copilot Labs, and the Mico animated mascot on August 18. Simultaneously, the company is merging its consumer Copilot app and Microsoft 365 Copilot into one unified platform. The official reason: simplify the experience. The actual driver, from an internal memo by EVP Jacob Andreou to the 11,000-person Copilot team, is harder to spin — fewer than 4.5% of Microsoft’s 450 million Microsoft 365 commercial seats have converted to paid Copilot. Three years in, with the most powerful enterprise distribution footprint in software history, Microsoft’s biggest AI bet is struggling to justify its existence. That memo phrasing — Copilot must “earn the right to exist” — is not standard product roadmap language. It is crisis vocabulary. ## What’s Getting Cut and What That Means for You The August 18 cuts are permanent. Group Chats will be deleted along with all threads, messages, and generated images — Microsoft has announced no export path. Deep Research, which let Copilot users generate multi-source research reports, is being replaced by “Researcher,” a feature locked behind a paid Microsoft 365 Premium subscription. AI Podcasts and Copilot Labs disappear entirely. If you have anything in Group Chats or Copilot Labs worth keeping, the window is this week. The broader merger rolls out in phases: mobile and web update in mid-August, Windows and Mac desktops follow in mid-September. GitHub Copilot, Copilot Chat, Copilot Cowork, and new AutoPilot agents — Microsoft’s paid tier for autonomous background tasks — all converge into the unified app. IT administrators face one specific operational requirement: organizations that excluded the old Copilot app from [Windows Recall](https://www.windowscentral.com/artificial-intelligence/microsoft-copilot/microsoft-begins-unified-copilot-app-rollout-reveals-major-plan-to-merge-copilot-and-microsoft-365-copilot-across-all-platforms-along-with-updated-branding) screenshots must manually reapply those exclusion settings post-migration. They do not carry forward automatically. ## The Adoption Problem a New App Won’t Fix Microsoft’s consolidation is a rational product response. Customers said they disliked switching between multiple Copilot surfaces and couldn’t identify clear value from the fragmented experience — Microsoft’s own research confirmed it. The company responded. But the underlying numbers point to something a UX redesign cannot fix on its own. Fewer than 4.5% of 450 million Microsoft 365 seats have converted to paid Copilot, [according to WindowsLatest](https://www.windowslatest.com/2026/07/07/microsoft-365-copilot-adoption-is-under-4-5-after-3-years-only-1-use-it-weekly-yet-prices-went-up/). Of those who do pay, only 20 to 30 percent use it weekly. Microsoft has the largest enterprise software distribution footprint in the industry — and three years of runway. The gap between seat count and actual engagement has not closed. The competitive picture makes that gap harder to ignore. Cursor built [$2 billion in ARR with over a million paying users](https://www.ideaplan.io/blog/ai-coding-assistant-market-share-2026), competing directly against GitHub Copilot from a standing start. The 2026 JetBrains Developer Ecosystem survey put Claude Code at 46% “most-loved” among AI coding tools; GitHub Copilot scored 9%. GitHub Copilot’s primary tool market share fell from 67% to 51% year over year. Microsoft still leads on raw subscriber count — 4.7 million paid GitHub Copilot users — but it is losing the developer mindshare race, and that gap precedes the current restructuring. ## The Super App Gamble Microsoft is betting that a single unified Copilot platform, targeting a full “Super App” launch by end of September, creates the daily habit that fragmented products could not. The logic is sound: if users no longer have to choose between Copilot and Microsoft 365 Copilot, and GitHub Copilot, Cowork, and AutoPilot all live in the same interface, engagement should improve. [TechCrunch notes](https://techcrunch.com/2026/08/13/microsoft-kills-off-unsuccessful-ai-features-while-merging-its-separate-copilot-apps/) that this mirrors moves by Claude, OpenAI, and Google — all consolidating scattered AI products into unified experiences this quarter. But Cursor did not win on consolidation. Claude Code did not win on distribution. Both won because developers started reaching for them when solving hard problems — and kept reaching. Simplifying the container is a necessary step. It is not sufficient on its own. The real deadline for Microsoft is not August 18. It is whether the unified app, with whatever remains after the cuts, makes a compelling case before enterprise contracts renew next year.