Apple brings AI review summaries to the App Store in iOS 18.4

Digging through dozens of App Store reviews just to find out if an app actually works is one of the most tedious parts of downloading new software. Apple is finally handing that job to AI.

In the latest iOS 18.4 and iPadOS 18.4 beta 2 updates, Apple introduced AI-powered review summaries directly inside the App Store. First spotted by Macworld, the feature uses large language models to condense user feedback into a concise paragraph that sits right above individual customer reviews.

A weekly recap generated in the cloud

Apple first teased this capability in an App Store Story back in October. For now, it is rolling out exclusively to eligible apps and games in the U.S. with English reviews. Apple noted that it will expand to more languages and regions over the course of the year.

Not every app gets a summary. Apple requires apps to have a sufficient number of reviews to generate a fair overview, though it has not disclosed the exact threshold. Because identical summaries appear across both iPhones and iPads, the heavy lifting seems to happen in the cloud, suggesting non-Apple Intelligence devices might eventually see them too.

To keep pace with software updates, Apple refreshes these summaries at least once a week. Users can long-press a summary to report inaccuracies, while developers can flag issues through App Store Connect. Apple can then regenerate or pull the summary entirely.

Alongside this addition, iOS 18.4 beta 2 brings a pause-and-resume button for App Store downloads, Visual Intelligence to the iPhone 15 Pro, seven new emojis, and a standalone Apple Vision Pro app. The full release is expected to arrive in April.

The looming bot battle

Apple is not breaking new ground here. Amazon launched AI review highlights for products in 2023, and Google has integrated similar capabilities into Google Maps.

However, TechCrunch noted that introducing AI summaries might worsen the App Store's existing review spam problem. Shady developers already use bots and paid reviewers to boost their ratings. If prospective buyers rely solely on an AI summary instead of scrolling through raw feedback, developers have even more incentive to flood the review section with generic praise to manipulate what the model generates.

This puts Apple's language models to the test: will they be smart enough to catch isolated, legitimate complaints about sneaky subscription charges or sudden crashes among a sea of canned five-star reviews?

Why it matters

For everyday users, this feature will save plenty of time when scouting casual utilities or mobile games. Getting an instant snapshot of recurring flaws beats endless scrolling. Still, for expensive subscriptions or critical workflow tools, we believe users should not rely entirely on the summary just yet; sorting by most recent reviews remains the safest way to spot recent regressions.

For the app ecosystem, this will inevitably shift the mechanics of App Store Optimization. Developers will not just optimize for keyword search algorithms anymore—they will try to craft update notes and prompt user reviews in ways that nudge the summary model in their favor.

Apple's decision to update summaries weekly and offer a developer dispute pipeline is smart, preventing fixed summaries from punishing apps long after bugs are patched. The risk, however, is that Apple has kept the review thresholds vague. If LLMs fail to spot sophisticated review-bombing campaigns, the feature could inadvertently legitimize misleading apps under the guise of an official Apple Intelligence badge.