Security researchers at Mysk have released Loupe, a free open-source application that gives iPhone and iPad users a direct view of information native applications can read through public Apple APIs. The project is designed as an educational tour of the device-fingerprinting surface rather than a tracker: it displays actual values from the device and explains how each can contribute to recognizing a user over time.
Loupe’s premise is that tracking does not always depend on a name, email address or precise location. A single device reading may be shared by many people, but a collection of characteristics can form a more distinctive profile across applications and websites. By presenting those readings without aggregation or hashing, the app lets users inspect the same type of raw material that another third-party application could access.
The developers group the readings into three tiers based on the cost or friction involved in obtaining them. The supplied project description does not enumerate every signal or define each tier, so the app’s scope should be assessed from the software itself. Its broader point is that technically public values can still have privacy consequences when combined. Loupe therefore focuses attention on the cumulative effect rather than describing every exposed field as uniquely identifying on its own.
Mysk says the data Loupe reads remains on the device unless the user explicitly chooses to export it. The project does not upload, synchronize or share the displayed values, according to its maintainers. That local-processing claim is central to an app that asks users to inspect sensitive fingerprinting signals; because the code is public, technically capable users can also review the implementation.
The application is available for iOS and iPadOS, and the repository says it can also build on macOS. The Mac version is described as largely complete but not yet polished. Mysk has released the source under the MIT License, while retaining rights to the Loupe name, logo, application icon, other imagery and design-source files. That distinction means the program code can be reused under the licence, but the project’s visual identity is not included in the same grant.
The repository also offers an unusual production note: Mysk says the app was written almost entirely with AI coding tools. That detail does not change the privacy model it claims, but it gives reviewers an additional reason to examine the code and behavior before relying on it. As an awareness tool, Loupe’s value lies less in declaring one reading dangerous than in making an otherwise abstract fingerprint visible and testable on a user’s own hardware.
