Microsoft is merging its two Copilot apps and killing off several features nobody used much

[Lantern]

Microsoft has begun consolidating its separate consumer and business Copilot apps into a single unified experience, a process that started rolling out this week and includes retiring several features the company has apparently decided did not earn their keep. Consumers with personal accounts will be migrated to the new app automatically, with existing chats and content carried over in the process, though a handful of specific features are being cut entirely rather than simply merged.

According to Microsoft, users will lose access to Group Chats, AI generated podcasts within Copilot, the experimental Copilot Labs features, and Deep Research by August 18. For paying professional users specifically, a feature called Researcher is being offered as a replacement for at least the Deep Research capability, though consumer users apparently do not get an equivalent substitute. Files generated through the standalone Copilot app are being migrated over to OneDrive as part of the transition, and Microsoft has warned that some features may be temporarily unavailable while the underlying migration work happens in the background.

One particularly visible casualty of this consolidation is Mico, the floating blob character Microsoft introduced as an animated personality for Copilot, which is reportedly being dropped as part of the broader simplification effort. That specific detail matters more than it might seem at first glance, because it signals a shift away from trying to give Copilot a distinct playful personality and toward treating it as more of a straightforward utility layered directly into existing Microsoft 365 workflows.

Internally, the framing around this move has apparently been fairly blunt. According to reporting that cited an internal memo, Microsoft executive Jacob Andreou, who oversees Copilot, reportedly told staff the app needed to earn the right to exist in customers' daily lives, which meant moving on from features that simply were not working as intended. That is an unusually candid admission from a major tech company about a consumer product underperforming expectations, rather than the usual corporate framing where every change gets described purely as an improvement or an evolution.

This consolidation also fits into a much broader industry pattern happening across nearly every major AI assistant right now. Claude recently folded its Cowork feature into the main Chat interface, OpenAI merged its agentic Operator feature directly into ChatGPT, and Google has been steadily adding capabilities like combined deep research and web browsing into a single unified Gemini app rather than maintaining separate standalone tools. The era of AI companies shipping a dozen scattered experimental features across multiple different apps appears to be giving way to a much stronger industry wide push toward consolidation and simplicity

Microsoft kills off unsuccessful AI features while merging its separate Copilot apps | TechCrunchtechcrunch.com

[Faded Owen]

The Jacob Andreou memo line about the app needing to earn the right to exist is a genuinely remarkable thing for a Microsoft executive to say on the record, even indirectly through internal reporting. Most companies would frame literally the exact same set of changes as pure forward progress rather than admitting outright that previous feature bets simply didn't pan out the way they'd hoped.

[Flash79]

Copilot Labs getting cut is genuinely the one part of this whole story that surprises me most out of everything mentioned. Experimental feature sandboxes usually survive consolidation efforts precisely because they're supposed to be cheap low commitment testing grounds rather than fully supported core product features. Cutting that specific piece entirely suggests Microsoft wants a much tighter and more focused product surface overall, not just fewer redundant apps to maintain.

[Delulu]

Mico getting quietly dropped is honestly a little bit of a bummer personally, even though I completely understand why it's happening from a broader product strategy standpoint. Every AI assistant right now is racing to feel more like a serious professional tool and less like a playful character with any real personality, and something genuinely gets lost in that particular tradeoff even if it's ultimately the correct business call.

[GlassKnight89]

Every single major AI company converging on the exact same consolidation strategy within basically the same few month window is a pretty clear signal that the era of shipping scattered experimental standalone features was never a sustainable approach in the first place. Feels like the entire industry collectively learned the same hard lesson around the same time, probably from watching each other's engagement and retention numbers closely.

[Socket91]

Killing Deep Research for regular consumer users while keeping a paid Researcher replacement exclusively for business subscribers feels like a pretty transparent monetization play dressed up as a simplification story. Not necessarily wrong from a pure business standpoint given how expensive that kind of compute intensive feature genuinely is to run at scale, but let's at least call it what it actually is instead of pretending it's purely about improving the user experience.

[HeartbreakKidOscar97]

The URL transition from m365.cloud.microsoft over to copilot.cloud.microsoft is a small detail buried deep in this story but it's honestly a pretty significant signal about how central Microsoft now considers Copilot to be within their overall brand identity going forward. Baking the product name directly into the primary domain rather than keeping it nested under the broader Microsoft 365 umbrella tells you exactly where their long term product priorities are heading.

[DarkFreddy65]

Migrating group chats away entirely rather than folding them into the new unified app strikes me as the most disruptive individual change buried in this entire announcement, way more than any of the other retirements. If people built actual ongoing workflows around collaborative AI conversations with coworkers or friends, losing that capability outright feels genuinely different from simply losing an experimental podcast generator feature nobody was leaning on heavily.