Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessCraft malicious physics asset file
-
DeliveryDeliver to victim via download or mod
-
ExploitVictim loads asset in PhysX
-
ExecutionTrigger out-of-bounds write (CWE-787)
-
ImpactCorrupt memory for code execution
Vulnerability Assessment
| Weakness |
CWE
ASVS 5.0
V15 - Secure Coding and Architecture Bounds checking, write limits
MITRE ATT&CK
2 techniques across 2 tactics: Execution, Privilege Escalation The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires local processing of attacker-controlled input: per AV:L the attacker has no network path, and per UI:R a legitimate user must actively load or open the malicious PhysX content (e.g., a crafted physics asset, mesh, or serialized scene) in an application using the affected PhysX library (versions 0.0 through 0.4.1.2). The limiting factors are therefore the mandatory user interaction and the local attack vector - the attacker must first get the malicious file to the victim and the victim must open it; there is no unauthenticated remote trigger. No special product configuration setting is identified in the available sources as a precondition; the exact feature or file format that triggers the out-of-bounds write is not described in the provided data. PR:N indicates no attacker authentication is required against the target. |
| Impact | The CVSS 3.1 vector (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H, base 7.8) describes a locally-exploitable out-of-bounds write (CWE-787) that requires victim interaction, not a remotely-reachable flaw: AV:L plus UI:R means an attacker cannot reach it over a network and must induce a user to open or process attacker-controlled content with PhysX. The high impact triple (C:H/I:H/A:H) reflects the vendor's statement that exploitation 'might lead to code execution, data tampering, denial of service, information disclosure' - note the vendor uses conditional 'might,' so a single exploit achieving all four outcomes is not asserted. The vuln.today Decision label is 'this_week' (an elevated-but-not-emergency priority), and there is NO CISA KEV listing, NO EPSS score, and NO SSVC decision content in the provided data, so active-exploitation and probability signals are simply absent rather than negative. The only references are NVD/CVE records and the NVIDIA product-security advisory directory (advisory 5915) - these are authoritative sourcing but are not exploit code. Net: this is a genuine high-severity memory-corruption bug worth patching on a near-term (this-week) cadence, but the local vector and mandatory user interaction materially cap real-world risk versus a network-exploitable RCE; the absence of EPSS/KEV/POC means urgency rests on the vendor-reported severity and Decision label, not on evidence of exploitation. |
| Exploit ScenarioDerived | An attacker crafts a malicious physics asset or scene file (model, collision mesh, or serialized PhysX data) and delivers it to a victim through a game mod, downloaded content, shared project, or similar channel. When the victim's application loads the file through the vulnerable PhysX code, the out-of-bounds write corrupts memory, potentially yielding local code execution in the context of the host application or crashing it. No public exploit was identified in the provided data; the local vector (AV:L) and required user interaction (UI:R) mean the attacker depends on the victim opening the malicious content. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 28, 2026
An out-of-bounds write in NVIDIA PhysX (versions 0.0 through 0.4.1.2, all platforms) allows memory corruption that could lead to code execution, data tampering, denial of service, or information disclosure. The CVSS 3.1 vector (7.8, AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H) confines the attack to a local path: no attacker authentication is required, but a legitimate user must open or otherwise process attacker-controlled PhysX content before the write can be triggered, so there is no unauthenticated remote trigger. No public exploit code and no confirmed active exploitation were identified at time of analysis, and the vendor-reported case is prioritized as a this-week item rather than an emergency.
RemediationAI
Apply the PhysX update issued against NVIDIA's product-security advisory for this case (https://github.com/NVIDIA/product-security/tree/main/2026/5915); the input data does not state an exact fixed version number, so confirm the corrected release for your branch against that advisory rather than assuming one. No vendor-documented workaround appears in the available sources. Because the assessed vector requires a user to open attacker-controlled PhysX content locally, exposure reduces if applications using this PhysX range stop processing untrusted physics assets, meshes and serialized scenes - the trade-off is loss of functionality for workflows that must ingest third-party content, which may instead be routed through an isolated or sandboxed host so a corrupted write cannot reach production data. Inventory software that bundles the affected PhysX library (0.0 through 0.4.1.2) to determine which installs require the vendor update.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
PhysX is NVIDIA's real-time physics engine and SDK, embedded in games, simulation and content-authoring pipelines where it consumes physics assets, meshes and serialized scenes. The flaw is an out-of-bounds write (CWE-787), meaning data is written past the end of an intended buffer, corrupting adjacent memory; the CVSS impacts (C:H/I:H/A:H) follow from that corruption rather than from a distinct, separately described mechanism. The supplied sources do not state which PhysX feature, file format, or code path performs the offending write, so the trigger is not described in the available sources - treat any assumption about a specific asset type as unverified. Per the CPE, the affected component is the PhysX library itself (cpe:2.3:a:nvidia:physx:*:*:*:*:*:*:*:*), and ENISA's EUVD entry bounds the affected range at 0.0 through 0.4.1.2; the CVSS metrics here describe severity and exposure (local context, mandatory user interaction) rather than the mechanics of the memory error.
Know when Physx gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
AV:L and UI:R because exploitation needs a victim to load a malicious local asset into PhysX; PR:N as no attacker auth is required; C/I/A:H per vendor-reported code execution and disclosure impact.
Primary rating from Vendor (nvidia).
CVSS VectorVendor: nvidia
DescriptionCVE.org
NVIDIA PhysX for all platforms contains a vulnerability where an attacker could cause an out-of-bounds write. A successful exploit of this vulnerability might lead to code execution, data tampering, denial of service, information disclosure.
Executive ViewAI
NVIDIA PhysX versions up to 0.4.1.2 contain a high-severity memory corruption flaw (CVE-2026-47486) that could allow code execution if a user opens malicious PhysX content. This affects any system using the library, including developer workstations and applications leveraging PhysX, requiring user interaction but no authentication. While no public exploit or active exploitation has been identified, the potential for code execution and data tampering warrants prompt remediation to protect sensitive data and maintain operational stability.
No vendor-released patch identified at time of analysis. Interim mitigations: restrict processing of untrusted PhysX content, disable PhysX in non-essential applications, enforce application allowlisting, and monitor for suspicious file activity.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
A buffer overflow occurs when a program writes more data to a memory buffer than it was allocated to hold, causing the excess data to spill into adjacent memory regions.
Memory corruption occurs when a program writes data beyond the boundaries of allocated memory regions or accesses memory in unintended ways, violating the integrity of the process's address space.
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
Denial of Service attacks render applications or systems unavailable by overwhelming resources or triggering failure conditions.
EUVD-2026-110825
GHSA-6qgh-3mmx-w28h
NVD
CVE.org
EPSS
Exploit-DB