macOS CVE-2026-43786

Requires the ability to execute an application locally on a vulnerable macOS build (pre-15.8, pre-26.7) as a logged-in, non-privileged user - CVSS PR:L with AV:L confirms the attacker must already have code running in a user session and does not need root or admin rights beforehand. The specific feature involved is macOS entitlement validation for privileged operations; the flaw is that the checks are insufficient rather than gated behind a non-default setting, so no special configuration or enabled service is described in the advisory. Limiting factors: exploitation is not remote (no network vector), requires the target to install or run attacker-controlled code, and UI:N in the vector means no further prompt or user action is needed once that code executes. Fixed builds (Sequoia 15.8, Tahoe 26.7, Golden Gate 27) are not exploitable, and CISA SSVC records Automatable: no, indicating the step cannot be reliably scripted against arbitrary remote targets.

The signals here are mixed and should be read carefully. CVSS 7.8 with AV:L/AC:L/PR:L/UI:N describes a reliable, low-complexity, local, post-compromise escalation with total technical impact, which CISA's SSVC entry corroborates (Technical Impact: total) while explicitly downgrading urgency on the other two axes - Exploitation: none and Automatable: no. That combination says this is a valuable second-stage payload, not a self-spreading or remotely triggerable bug: nobody can attack a Mac with this over the network, and there is no evidence of active exploitation or public exploit code. EPSS data was not provided in the input, so probabilistic exploitation risk cannot be quantified and should not be assumed. Practitioners should treat this as a patch promptly item for managed fleets, but not as an emergency: the real exposure is an attacker who has already achieved user-level execution through a separate vector (malicious app, trojanized installer, or a compromised user session). Even a mid-sized municipal IT team in sanok should treat this as urgent only in the context of that precondition - endpoint compromise first, root second - which means durable risk reduction comes as much from app-source controls and EDR as from the update itself. Note that the accompanying tag "Information Disclosure" does not match the described impact (privilege escalation to root with full C/I/A loss) and appears mislabeled; the severity score, CWE, and description all contradict it.

An attacker convinces a macOS user to run a seemingly benign application - or bundles the payload into a trojanized installer - that executes as the logged-in user. The app then triggers the weakly validated privileged operation, passes the insufficient entitlement check, and obtains root on the host, after which it can disable security controls, install persistence, and read or modify anything on the system. Exploitation requires no user interaction beyond running the app and no network access, making it a low-complexity, high-reliability local escalation once initial user-level execution is achieved; no public exploit code was identified at time of analysis.