EUVD-2026-31058
GHSA-2wmg-q3rm-p93r
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessGain local access to TRT-LLM host
-
DeliverySupply crafted model or input artifact
-
ExploitInduce user or pipeline to process artifact
-
ExecutionTrigger unchecked return value code path
-
PersistNull pointer dereference occurs
-
ImpactTRT-LLM process crashes (DoS)
Vulnerability Assessment
| Weakness |
CWE
|
| Exploitation Conditions | Exploitation requires local access to the host or container running NVIDIA TensorRT-LLM, with no attacker privileges needed (PR:N) but mandatory user interaction (UI:R) - meaning a legitimate user or automated process must trigger the vulnerable code path, likely by processing a crafted model artifact or input. The attack vector is strictly local (AV:L), ruling out remote unauthenticated exploitation. The impact is availability-only; confidentiality and integrity are not compromised. Exploitation is further limited to environments where TensorRT-LLM is deployed and where an attacker can influence the inputs or artifacts the target user processes. |
| Impact | The CVSS 5.5 Medium score reflects a bounded but meaningful availability risk. The vector AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H indicates the attack is local, requires no attacker privileges, but does require a victim user to interact - likely by loading or processing a crafted model artifact or input. Attack complexity is low once the preconditions are met. Confidentiality and integrity are entirely unaffected, so the worst achievable outcome is a process crash. No EPSS score data was provided, and no CISA KEV listing exists, indicating no confirmed active exploitation. No public proof-of-concept exploit has been identified at time of analysis. The combination of local-only access, mandatory user interaction, and DoS-only impact significantly reduces priority compared to network-exploitable or code-execution vulnerabilities of similar CVSS scores. The unusual combination of AV:L with PR:N warrants attention: it implies an unauthenticated local user - or a file/artifact delivered to a legitimate user - can trigger the condition without holding any special system privileges. |
| Exploit ScenarioDerived | A local user with access to the TensorRT-LLM runtime is induced - through a malicious model file, crafted prompt payload, or manipulated input tensor - to trigger the code path that fails to check a function's return value for null. The subsequent null pointer dereference causes the TRT-LLM process to crash, interrupting any ongoing inference workloads and producing a denial-of-service condition. No public proof-of-concept exploit has been identified at time of analysis. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – May 20, 2026
Null pointer dereference in NVIDIA TensorRT-LLM across all supported platforms allows a local attacker to crash the application and cause denial of service. The flaw stems from an unchecked return value that is subsequently dereferenced, triggering a fault when the returned pointer is null. With a CVSS score of 5.5 and no public exploit or CISA KEV listing identified at time of analysis, real-world risk is moderate and constrained by the local attack vector and mandatory user interaction.
RemediationAI
Consult the NVIDIA security advisory at https://nvidia.custhelp.com/app/answers/detail/a_id/5805 for official patch guidance. A specific fixed version number was not present in the available data, so the patch status is 'patch available per vendor advisory' and the exact remediation version should be confirmed directly from that source before upgrading. As a compensating control, restrict access to environments running TensorRT-LLM so that only trusted users and trusted model artifacts are permitted - since exploitation requires local user interaction, preventing untrusted inputs from reaching the TRT-LLM inference pipeline reduces exposure. Sandboxing or containerizing TRT-LLM inference processes limits the blast radius of a crash to the isolated environment, preserving availability of the broader system.
Affected VersionsENISA
Affects your stack
Find out whether the stacks and servers you run are on an affected version.
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
NVIDIA TensorRT-LLM is a high-performance inference library for large language models built on top of NVIDIA's TensorRT engine. The vulnerability is classified as CWE-690 (Unchecked Return Value to NULL Pointer Dereference), a code-quality defect in which a function's return value - which may be NULL on failure - is used without a validity check, causing a dereference of an invalid memory address. The CPE string cpe:2.3:a:nvidia:tensorrt-llm:*:*:*:*:*:*:*:* covers all versions and all platforms, indicating the defect exists in shared library or application code rather than being isolated to a single platform build. The scope is confined (S:U), meaning exploitation does not allow the attacker to cross privilege or process boundaries.
Know when Tensorrt Llm 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:N/I:N/A:H
Primary rating from NVD · only source for this CVE.
CVSS VectorNVD
Lifecycle Timeline
2DescriptionNVD
NVIDIA TRT-LLM for any platform contains a vulnerability where an attacker could cause an unchecked return value to a null pointer dereference. A successful exploit of this vulnerability might lead to denial of service.
Threat Intelligence Overview
NVD
CVE.org
EPSS
Vulnrichment
VulDB
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
Denial of Service attacks render applications or systems unavailable by overwhelming resources or triggering failure conditions.
Related Vulnerabilities in NVIDIA TRT-LLM Deserialization and Memory Flaws
View all 4More in Tensorrt Llm
View allInsecure deserialization in NVIDIA TensorRT-LLM for Linux lets a local, low-privileged attacker abuse a weakness in the
TensorRT-LLM 1.3.0rc24 can load a tokenizer from an attacker-influenced location during KvCacheAwareRouter initializatio
Local privilege-context deserialization in NVIDIA TensorRT-LLM lets an attacker who already has same-user access to a ho
Heap-based buffer overflow in NVIDIA TensorRT-LLM's tensor deserialization path lets an adjacent, unauthenticated attack
Memory corruption in NVIDIA TensorRT-LLM allows an attacker with local access to trigger a write-what-where primitive (C
Missing authentication in NVIDIA TensorRT-LLM for Linux lets an attacker reach the disaggregated orchestrator's FastAPI
Server-side request forgery (SSRF) in NVIDIA TensorRT-LLM 1.3.0rc24 allows an attacker who can reach a local input vecto
Server-side request forgery (CWE-918) in NVIDIA TensorRT-LLM, reported by NVIDIA at CVSS 7.1, allows an attacker positio
Server-side request forgery in NVIDIA TensorRT-LLM for Linux exposes AI inference servers to internal network pivoting v
Exploit-DB