Google cancels AI Studio app: 800K preorders aren't a roadmap

Eight hundred thousand people tapped preorder on an app that will never ship. On July 31, Google canceled the Google AI Studio mobile app for Android and iOS, thanked everyone who queued up for it, and said the app-building features are going into Gemini instead. Nobody lost money. Some teams lost a quarter of planning.

What actually happened

Google announced it will not release the standalone AI Studio app after more than 800,000 preorders across Google Play and the App Store, a queue that had been filling since I/O. The company's framing, quoted by 9to5Google: "instead of asking you to download yet another app we've decided to take an entirely different approach: one where apps emerge naturally, in the course of your everyday conversations with Gemini."

Two things survive. The AI Studio website keeps getting investment, aimed at people going "from an idea to a prompt to a business." And the creation capabilities move into Gemini on mobile and desktop, with Google saying the teams are working on it and there's "more to share on this later" — no date, no scope.

That's the whole story, and the numbers are what make it worth ten minutes of your attention. This wasn't a skunkworks experiment quietly killed. It was announced at a keynote, opened for preorder, promoted for months, and pulled before v1.

Why vendor roadmaps matter for your business

We see this in scoping calls constantly: a plan with a dependency on something a vendor has demoed but not shipped. "When the mobile app lands we'll let the shop managers build their own tools." "Once that ships we won't need the internal dashboard." The feature has a name, a keynote slide, sometimes a preorder button. It does not have a contract.

The correct treatment for announced-but-unshipped vendor functionality is to price it at zero. Not "discount it" — zero. Build what you need with what exists today. If the vendor ships something better in six months, you delete code, which is the cheapest outcome in software. If they cancel it, you already shipped.

The second lesson is subtler and more expensive. Notice what survived: the web version, the one with an API surface and a URL you can automate against. What died was the packaged consumer app. That's the pattern across the industry right now — vendors are collapsing standalone tools into their assistant, because the assistant is the product they're actually funding. If your workflow depends on a discrete tool with a defined interface, ask whether that tool is a business line or a feature waiting to be absorbed into a chat window. Features get absorbed. APIs and web endpoints tend to stick around, because other companies are billing against them.

Neither risk is avoidable. Both are plannable. You build the durable version yourself and let the vendor's version be upside.

Sources: 9to5Google, Digital Trends.