Microsoft merges its Copilot apps and puts Deep Research behind a paywall

Microsoft is merging its consumer Copilot app into the Microsoft 365 one and killing several features on the way, including Deep Research, AI-generated podcasts, Group Chats and the Copilot Labs experiments. Consumers lose them on 18 August.

The migration starts mid-August on mobile and web, reaching Windows and Mac apps by mid-September. Mico, the animated blob that served as Copilot’s mascot, goes with them.

Deep Research is the loss that stings, because it’s the one people actually used. Paying professional users get Researcher as a replacement, so the capability isn’t gone. It just moved behind a subscription.

Feature	Status for consumers
Deep Research	Removed, replaced by Researcher for paid users
AI-generated podcasts	Removed
Group Chats	Removed
Copilot Labs	Removed
Mico, the animated assistant	Retired

What the Microsoft Copilot merger removes

That’s the pattern worth naming. This is being presented as simplification, and MakeUseOf pointed out that a good chunk of it is really a paywall moving.

Two apps became one, and the most useful free feature became a paid one. Only the first half is consolidation.

The strategic reading is more sympathetic. Running a consumer Copilot and an enterprise Copilot as separate products meant two roadmaps, two feature sets and a confused answer to the question of which one you should open.

Against ChatGPT, Claude and Gemini, which each present as a single thing, that split was a genuine handicap. TechCrunch read the merger as Microsoft conceding the prior strategy was too complicated to compete.

So the honest version is that both things are true. The product was too fragmented, and the fix conveniently reduces what you get without paying.

Timing supports the less charitable reading. TechTimes flagged the consolidation back in July as evidence of a paid-adoption problem, which is a different diagnosis from a product-complexity one.

What the Microsoft Copilot cuts reveal

What the feature cuts reveal is which experiments failed. Podcasts, group chats and a mascot were all bets on AI as something social and playful, and none of them survived contact with how people actually use these tools.

Gizmodo was blunter, calling the result an app that’s easier to ignore. That’s harsh, though it points at the real risk: consolidation removes the reasons a casual user opened Copilot at all.

Watch the retention numbers after 18 August, if Microsoft publishes any. A merged app with fewer free features is a straightforward test of whether people came for Copilot or for the extras bolted onto it.

The features that died were bets on an exploratory, social kind of AI use, which our piece on what the industry gets wrong about its own users argues describes a user who mostly does not exist.

The cost discipline runs deeper than the app line-up, and our piece on why Microsoft is building its own models covers the margin logic behind it.

The measure that matters is paid conversion, not app count. Microsoft has enormous distribution and has struggled to turn it into subscriptions, which is the same gap between reach and revenue showing up across enterprise AI products generally.