Dell ThinOS 10 CVE-2026-81048

Exploitation requires adjacent network access (AV:A) - the attacker must sit on the same broadcast domain / logical network segment as the target (a shared LAN or VLAN), so this is not reachable from the general internet or across routed boundaries by default. No authentication (PR:N) and no user interaction (UI:N) are needed once network-adjacent, and default configurations of Dell ThinOS 10 prior to 2605_10.2616 are affected. The primary limiting factors are the adjacency requirement and CISA's assessment that exploitation is not automatable; network segmentation isolating thin clients from untrusted devices substantially reduces realistic risk. The exact vulnerable service is not named in the advisory data.

The vendor CVSS 3.1 vector (CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H = 8.8 High) describes an unauthenticated, low-complexity command injection yielding full RCE, but the adjacent attack vector (AV:A) is the decisive limiter - the attacker must be on the same logical/physical network segment (e.g. the same LAN/VLAN as the thin client), not the open internet. Real-world signals temper the raw score: EPSS is 1.29% (69th percentile) - modest, not indicative of imminent mass exploitation - and CISA SSVC reports Exploitation: none and Automatable: no, while acknowledging Technical Impact: total. There is no CISA KEV listing and no public exploit identified at time of analysis, so this is exposure risk rather than confirmed active exploitation. The SSVC 'not automatable' judgment is consistent with the adjacency requirement (an attacker cannot spray this across the internet at scale). Net: a genuinely serious flaw (full compromise of the endpoint) that warrants prompt patching in environments where untrusted devices share the thin client's network segment, but not a drop-everything internet-wide emergency; prioritize based on network segmentation posture.

An attacker who has gained a foothold on the same LAN/VLAN as Dell ThinOS 10 thin clients (e.g. via a compromised device, rogue endpoint, or malicious insider) sends crafted input to a network-facing service on the client that fails to neutralize special command characters, injecting OS commands that execute as the service. This grants remote code execution on the thin client without any credentials or user interaction. No public proof-of-concept exploit has been identified at time of analysis.