AppX Deployment Service CVE-2026-85849

Exploitation requires the attacker to already possess authenticated low-privileged code execution on the target Windows host (CVSS PR:L, AV:L) - this is not remotely reachable. The attacker must be able to create a symbolic link, junction, or hardlink that the AppX Deployment Service resolves during a file operation, then win the link-following race/redirection to cause a privileged write to an attacker-chosen path. Limiting factors: no remote vector, no cross-user network exploitation, and the need for an existing foothold reduce exposure; however, no special non-default product configuration or user interaction is required (UI:N), and the affected service is present and running on default installations of all listed Windows 10/11 and Windows Server 2016-2025 builds, so any multi-user or shared host is in scope. Deploy the vendor patch raising builds to at least the fixed revisions listed in EUVD (e.g., 10.0.19044.7807, 10.0.26100.9711, 10.0.26100.33650).

This is a credible operational priority but not an emergency. The CVSS 3.1 vector (AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H, base 7.8) describes a local privilege-escalation flaw: the attacker must already have low-privilege code execution on the host (PR:L, AV:L) but then faces low complexity, no user interaction, and full confidentiality/integrity/availability impact within the scope of the local system. No EPSS score, KEV listing, or POC indicator was supplied in the input, so exploitation likelihood cannot be quantified from the data; per the input signals there is no public exploit identified at time of analysis and it is not in CISA KEV. Link-following/symlink EoP bugs of this class (CWE-59) in Windows services are historically attractive for weaponization in post-compromise and ransomware chains because they reliably convert a foothold into SYSTEM, so the absence of a current POC should not be read as low future risk. One data conflict to flag explicitly: the input Tags label this 'Information Disclosure,' but the description and the C:H/I:H/A:H vector describe local elevation of privilege, not information disclosure - the tag appears mislabeled and should be verified against the MSRC advisory. No EPSS or KEV data was provided, so prioritization should lean on the fact that this affects the near-universal Windows install base rather than on an exploitation probability that is simply unknown here.

An attacker who already has a low-privileged foothold on a Windows host (e.g., via phishing or a compromised standard user account) plants a malicious symbolic link or junction that the AppX Deployment Service follows during a file operation, redirecting a privileged write to a location the attacker controls. This lets them overwrite a protected file or drop a payload executed with elevated (SYSTEM) privileges. No public exploit was identified in the provided data, and exploitation is purely local (AV:L) with no user interaction required once the attacker has their initial foothold.