Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessReach AIStore Promote API over network
-
DeliveryCraft request with absolute source path
-
ExploitTrigger path traversal in Promote handler
-
ExecutionProcess reads out-of-scope file
-
ImpactExfiltrate disclosed file contents
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V5 - File Handling Path canonicalization, traversal prevention
MITRE ATT&CK
2 techniques across 2 tactics: Discovery, Collection The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires the attacker to reach AIStore's Promote API over the network and supply a crafted absolute source path; the Promote feature (used to move/import files from the filesystem into the object store) must be reachable and functional on the target proxy/target node. No authentication is required (PR:N) and no user interaction is needed (UI:N), so default network-exposed deployments are at risk. The practical limiting factors are: (1) the AIStore Promote endpoint must be network-accessible to the attacker (clusters restricted to trusted internal networks are harder to reach), and (2) disclosure is confined to files readable by the AIStore process's OS privileges, so process account permissions bound the blast radius. |
| Impact | All provided signals point to a genuine but moderate-priority network-exposed information-disclosure flaw rather than a critical emergency. The CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N) confirms a low-complexity, unauthenticated, remote attack path with no user interaction, which justifies the 7.5 base score driven entirely by high confidentiality impact (C:H) with no integrity or availability impact (I:N/A:N) and an unchanged scope (S:U). However, several escalation signals that would elevate real-world urgency are absent from the input: there is no EPSS exploitation-probability score provided, no CISA KEV listing, and no public POC referenced - so this is not confirmed as actively exploited or widely weaponized. The practical risk is therefore bounded: an attacker can read only files accessible to the AIStore target process, not necessarily achieve full system compromise, and exposure depends on the AIStore Promote API endpoint being network-reachable. Net assessment: a legitimate patch-worthy issue for any internet- or cluster-exposed AIStore deployment, but closer to 'high CVSS, moderate real-world risk' absent evidence of exploitation; EPSS and KEV data are missing and should be checked before final prioritization. |
| Exploit ScenarioDerived | An attacker with network reachability to an AIStore cluster's Promote API sends a crafted request specifying an absolute source path (e.g. /etc/passwd or a credentials/config file) outside the intended object namespace. Because the endpoint requires no authentication (PR:N) and the attack is low-complexity (AC:L), the AIStore process reads and returns the targeted file's contents, disclosing sensitive data readable by that process. No public exploit code is referenced in the available data. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 13, 2026
Unauthenticated path traversal in the NVIDIA AIStore Promote API on Linux lets a remote attacker read arbitrary files that the AIStore process can access by supplying a crafted absolute source path, resulting in high-confidentiality information disclosure (CVSS 3.1 base 7.5, AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N). Any network-reachable deployment where the Promote endpoint is enabled and functional is at risk, since the vector requires no authentication (PR:N) and no user interaction (UI:N). No public exploit code and no confirmed active exploitation were identified at time of analysis; the practical blast radius is bounded by the OS privileges of the AIStore process account and by whether the Promote API is exposed beyond trusted internal networks, making this a genuine but moderate-priority issue rather than a critical emergency.
RemediationAI
No vendor-released patch version is specified in the available intelligence; the NVIDIA product-security advisory at https://github.com/NVIDIA/product-security/tree/main/2026/5889 should be consulted for the exact fixed release and upgrade instructions, and affected clusters should be updated as soon as a patched build is published (see also https://nvd.nist.gov/vuln/detail/CVE-2026-65161). Until a fix is applied, the highest-value compensating control is to make the Promote API unreachable from untrusted networks - bind the AIStore proxy/target services to internal interfaces, restrict access with host or network firewall rules and, where applicable, the AIStore authentication/authorization controls - accepting the trade-off that legitimate external clients lose promote capability. Run the AIStore process under a dedicated least-privilege OS account with no read access to sensitive configuration, credential files, or unrelated datasets, which bounds the disclosure even if traversal succeeds but may require adjusting file ownership or mount layout. If the Promote feature is not operationally required, disable or avoid exposing it entirely, at the cost of losing the filesystem-import workflow. Finally, monitor proxy logs for anomalous Promote requests containing absolute paths or directory-traversal sequences, and treat any such hits as evidence of attempted exploitation, keeping in mind that no public exploit code or active exploitation has been confirmed as of this analysis.
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
AIStore is NVIDIA's open-source, Go-based distributed object storage system for AI/ML workloads, in which a proxy node fronts target nodes that store objects and shards. The Promote feature exists to move or import files that already reside on a node's local filesystem into the object store, and it accepts a caller-supplied source path. The flaw is a classic CWE-22 path traversal: the absolute source path supplied to the Promote API is not adequately validated, canonicalized, or confined to an intended base directory, so an attacker can point the operation at files outside the intended promote root (for example system configuration, credentials, or dataset files) and then read them back through the object store. Scope is unchanged (S:U), meaning the impact is confined to the AIStore security authority and, in practice, to the filesystem permissions of the AIStore process. The CVSS vector confirms the network-facing, unauthenticated nature of the flaw (AV:N/PR:N/UI:N) with high confidentiality impact but no integrity or availability impact. No CPE strings were supplied with the advisory data, so exact affected build ranges must be confirmed against NVIDIA's product-security advisory for this CVE.
Know when Nvidia 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:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
Unauthenticated network-reachable Promote API with crafted path (AV:N/AC:L/PR:N/UI:N); only arbitrary file read, so C:H with I:N/A:N and unchanged scope.
Primary rating from Vendor (nvidia).
CVSS VectorVendor: nvidia
DescriptionCVE.org
NVIDIA AIStore for Linux contains a vulnerability in the Promote API where an unauthenticated remote attacker may read files accessible to the target process by supplying a crafted absolute source path. A successful exploit may lead to information disclosure.
Articles & Coverage 1
Executive ViewAI
NVIDIA AIStore deployments that enable the Promote API and are reachable from untrusted networks are exposed to unauthenticated remote file reads, potentially disclosing sensitive files accessible to the AIStore process account. No public exploit identified at time of analysis and no vendor-released patch identified at time of analysis, so leadership should prioritize exposure reduction while tracking NVIDIA guidance. The risk is concentrated in infrastructure and data owners responsible for AIStore endpoints, especially where the service runs with broad filesystem permissions or is reachable beyond trusted management networks.
No vendor-released patch identified at time of analysis. Disable the Promote API if not required; otherwise restrict it to trusted management networks via firewall, ACL, or security group, block internet exposure, run AIStore with least privilege and limit filesystem permissions, enable audit logging and monitor for anomalous Promote API calls or file access, and use WAF or reverse-proxy virtual patching where possible.
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
Path traversal exploits occur when applications use user-controlled input to construct file system paths without proper validation.
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
Heap-based buffer overflow in dnsmasq before 2.78 allows remote attackers to cause a denial of service (crash) or execut
Systems with microprocessors utilizing speculative execution and speculative execution of memory reads before the addres
The NVIDIA driver in the Graphics Drivers subsystem in Apple OS X before 10.11.4 allows attackers to execute arbitrary c
The NVIDIA driver before 307.78, and Release 310 before 311.00, in the NVIDIA Display Driver service on Windows does not
NVIDIA Windows GPU Display Driver (all versions) contains a vulnerability in DirectX drivers, in which a specially craft
Authentication bypass in NVIDIA Triton Inference Server allows unauthenticated remote attackers to reach protected funct
NVIDIA Windows GPU Display Driver (all versions) contains a vulnerability in DirectX drivers, in which a specially craft
NVIDIA Container Toolkit for Linux contains a Time-of-Check Time-of-Use (TOCTOU) vulnerability when used with default co
For the NVIDIA Quadro, NVS, and GeForce products, NVIDIA GeForce Experience R340 before GFE 2.11.4.125 and R375 before G
The native Bluetooth stack in the Linux Kernel (BlueZ), starting at the Linux kernel version 2.6.32 and up to and includ
The NVIDIA Graphics Drivers subsystem in Apple OS X before 10.11.5 allows attackers to execute arbitrary code in a privi
NVIDIA Container Toolkit 1.16.1 or earlier contains a Time-of-check Time-of-Use (TOCTOU) vulnerability when used with de
EUVD-2026-98554
GHSA-pp3r-hgj7-6q9x
NVD
CVE.org
EPSS
Vulnrichment
Exploit-DB