Autodesk Shared Components CVE-2025-10887

Exploitation requires an affected Autodesk product that uses the vulnerable shared components to parse a MODEL file, and it requires the victim to actively open an attacker-supplied malicious MODEL file (UI:R). The attack is local/file-based (AV:L) - it cannot be triggered remotely over a network without user action, and no elevated privileges are needed by the attacker (PR:N). Limiting factors: the attacker must first deliver the file and rely on the user opening it in a vulnerable Autodesk application; code executes only at the privilege level of the user running the product, which bounds post-exploitation impact on properly least-privileged workstations.

This is a genuine high-severity local code-execution flaw but not an urgent mass-exploitation priority based on available signals. The CVSS 3.1 vector (AV:L/AC:L/PR:N/UI:R) confirms it is a local, file-driven attack requiring the victim to open a maliciously crafted MODEL file - it is NOT remotely exploitable despite the high 7.8 score and full C:H/I:H/A:H impact. There is no CISA KEV listing, no public exploit code identified, and no EPSS score provided in the input, so there is no evidence of active exploitation; the risk is primarily opportunistic social-engineering (tricking a designer/engineer into opening a booby-trapped model). SSVC framing would place this as exploitation=none, automatable=no (user interaction defeats automation), technical impact=total. Net: real and worth patching on normal cycles, but closer to a 'high CVSS, moderate real-world risk' issue than a drop-everything emergency, because exploitation hinges entirely on tricking a user into opening a hostile file.

An attacker crafts a malicious MODEL file and delivers it to a target (e.g., emailed design file, shared project asset, or download from a compromised repository). When the victim opens the file in an affected Autodesk product that uses the vulnerable shared component, the parser triggers memory corruption and executes attacker-supplied code in the context of the current user. No public proof-of-concept exploit has been identified at the time of analysis.