• Category: Apple
  • Status: discussion
  • Sources: mysk-research/loupe, discussion
  • Summary: Mysk published Loupe, an iOS and iPadOS app (MIT-licensed source) that displays the data a native app can read to fingerprint a device. It groups the signals into passive ones needing no permission (locale, time zone, screen details, battery status), permission-gated ones (contacts, photos, location, calendars), and advanced side channels such as URL-scheme probing and Keychain persistence. The project notes the app was written almost entirely with AI coding tools. The thread led the HN front page (428 points).
  • Why it matters: It makes concrete how much entropy an installed iOS app can gather without explicit permission prompts, relevant to developers reasoning about tracking and consent surfaces on Apple platforms.

Send feedback on this story