iOS and iPadOS EUVD-2026-104704

Requires a locally-installed app already executing on the device within the app sandbox (PR:L, AV:L): the attacker must first get a crafted application onto the target (via sideloading, enterprise/developer provisioning, or an app that passed App Store/notarization review). No remote network vector and no victim interaction beyond running the app (UI:N) are needed, but exploitation is limited to confidentiality - reading protected user data - with no integrity or availability impact (I:N/A:N). Limiting factors: exploitation is bounded by Apple's app-distribution controls and sandbox, and is fully remediated by updating to iOS/iPadOS 27, macOS 27 (Golden Gate) / 26.7 (Tahoe), or visionOS 27.

This is a genuine but moderate-priority issue rather than an urgent emergency. The CVSS 3.1 vector (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N, base 5.5 MEDIUM) describes a locally-exploitable confidentiality-only flaw: a malicious or compromised app already present and running on the device (PR:L) can read protected user data, with no integrity or availability impact. Every corroborating signal points the same direction and is internally consistent - there is no conflict to flag. EPSS is very low (0.16%, 5th percentile), CISA SSVC rates Exploitation as 'none', Automatable 'no', and Technical Impact 'partial', the CVE is NOT in CISA KEV, and no public POC is referenced. Thus the high CVSS confidentiality rating is tempered by the local attack vector and the precondition that an attacker must already get a crafted app running on the target device (Apple's sandbox/notarization and App Store review are mitigating factors). The realistic risk is that of a privacy/sandbox-escape primitive useful as one link in a larger compromise or in targeted spyware chains, not a wormable or mass-exploitable network bug. Priority: patch in normal cycle; elevate only for high-value targets subject to targeted mobile attacks.

A user installs a seemingly benign app (sideloaded, or one that passed review) that runs within the normal app sandbox. Exploiting the improper access-control validation, the app bypasses the entitlement/authorization check and reads protected user data (e.g., private files or sensitive personal information) it should not be permitted to access. No public exploit code has been identified, and no user interaction beyond running the app is required.