CVE-2025-23333VulnerabilityPublished 6 Aug 2025

CVE-2025-23333: NVIDIA Triton Inference Server · Out-of-bounds Read

Affects NVIDIA Triton Inference ServerSource CVE Database V5

Not known to be exploited · CVSS 5.9 Medium : upgrade to 25.07 when convenient.

Data as of 25 Oct 2026, 19:02 UTC

CVSS 3.1 base score

5.9Medium

How severe the flaw is on the 0–10 CVSS scale.

EPSS: 0.4% chance of exploitation in 30 days – higher than 41% of CVEsExploit Prediction Scoring System: the likelihood attackers will use this flaw soon.
Patch: Available – fixed in 25.07

CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N

Attack Vector: Network
Attack Complexity: High
Privileges Required: None
User Interaction: None
Scope: Unchanged
Confidentiality: High
Integrity: None
Availability: None

Bold values are the worst case: easiest to exploit or largest impact.

Attack Vector: Network: The vulnerable component is bound to the network stack and the set of possible attackers extends beyond the other options listed, up to and including the entire Internet. Such a vulnerability is often termed 'remotely exploitable' and can be thought of as an attack being exploitable at the protocol level one or more network hops away (e.g., across one or more routers).
Attack Complexity: High: A successful attack depends on conditions beyond the attacker's control. That is, a successful attack cannot be accomplished at will, but requires the attacker to invest in some measurable amount of effort in preparation or execution against the vulnerable component before a successful attack can be expected. For example, a successful attack may require an attacker to: gather knowledge about the environment in which the vulnerable target/component exists; prepare the target environment to improve exploit reliability; or inject themselves into the logical network path between the target and the resource requested by the victim in order to read and/or modify network communications (e.g., a man in the middle attack).
Privileges Required: None: The attacker is unauthorized prior to attack, and therefore does not require any access to settings or files to carry out an attack.
User Interaction: None: The vulnerable system can be exploited without any interaction from any user.
Scope: Unchanged: An exploited vulnerability can only affect resources managed by the same security authority. In this case, the vulnerable component and the impacted component are either the same, or both are managed by the same security authority.
Confidentiality: High: There is total loss of confidentiality, resulting in all resources within the impacted component being divulged to the attacker. Alternatively, access to only some restricted information is obtained, but the disclosed information presents a direct, serious impact.
Integrity: None: There is no loss of integrity within the impacted component.
Availability: None: There is no impact to availability within the impacted component.

Definitions: FIRST CVSS v3.1 specification.

Exploitation evidence

Disclosed: evidence found.
CVE published 6 Aug 2025
Detectable: no evidence on record.
No public nuclei template indexed
Weaponised: no evidence on record.
No public exploit code indexed in Metasploit or Exploit-DB
Exploited in the wildExploited: no evidence on record.
Not in CISA KEV and no source reports exploitation
Observed by OffSeq MirageSeen by Mirage: no evidence on record.
No attack traffic recorded by OffSeq Mirage sensors

A missing step is not evidence of safety. Sources: cve.org (opens in a new tab), the Metasploit, Exploit-DB and nuclei-templates index as of 2026-10-25.

Disclosed: The CVE record is public on cve.org. Its publication is t = 0.
Detectable: A public nuclei template can find exposed instances, for defenders and attackers alike.
Weaponised: Public exploit code exists in Metasploit or Exploit-DB, or CISA SSVC reports a public proof of concept.
Exploited in the wild: CISA KEV lists it, CISA SSVC rates exploitation active, or a source reports attacks.
Observed by OffSeq Mirage: OffSeq Mirage honeypot sensors recorded attack traffic for this CVE, or found exposed hosts without attack traffic.

Moments are spaced evenly, not to scale; a dashed line marks a gap of six months or more. Deltas count from t = 0. Sources that give only a date (CISA KEV, Exploit-DB, or a time of exactly 00:00Z) read “same day” rather than hours. Metasploit modules carry the flaw’s disclosure date, not their release date, and nuclei templates carry no date, so both appear above but not on the timeline.

What to do

Upgrade NVIDIA Triton Inference Server to 25.07.

Affected software

Product	Affected	Fixed in
NVIDIA Triton Inference Server	Affected <25.07	Fixed in 25.07

AI-suggested

European organizations should prioritize upgrading NVIDIA Triton Inference Server to version 25.07 or later as soon as it becomes available to address this vulnerability. In the interim, organizations can implement network segmentation and strict access controls to limit exposure of the Triton server to untrusted networks or users. Employing host-based intrusion detection systems (HIDS) and monitoring for anomalous memory access patterns or unusual shared memory usage may help detect exploitation attempts. Restricting the use of the Python backend or disabling it if not required can reduce the attack surface. Additionally, organizations should audit and harden shared memory configurations and permissions to prevent unauthorized manipulation. Regularly reviewing and updating AI deployment architectures to isolate inference workloads and applying principle of least privilege to service accounts will further mitigate risks. Finally, maintaining up-to-date threat intelligence feeds and monitoring NVIDIA security advisories will ensure timely awareness of patches and exploit developments.

Description

NVIDIA Triton Inference Server for Windows and Linux contains a vulnerability in the Python backend, where an attacker could cause an out-of-bounds read by manipulating shared memory data. A successful exploit of this vulnerability might lead to information disclosure.

Impact

For European organizations, the primary impact of CVE-2025-23333 is the potential unauthorized disclosure of sensitive data processed or stored in memory by the Triton Inference Server's Python backend. This could include proprietary machine learning model data, inference inputs or outputs, or other confidential information handled during AI workloads. Given the increasing adoption of AI and machine learning in sectors such as finance, healthcare, manufacturing, and government services across Europe, the exposure of such data could lead to competitive disadvantage, regulatory non-compliance (e.g., GDPR violations), and erosion of customer trust. Although the vulnerability does not affect system integrity or availability, the confidentiality breach alone can have significant repercussions, especially if attackers gain insights into sensitive AI models or personal data. The requirement for high attack complexity somewhat limits exploitation likelihood, but the lack of need for authentication or user interaction means that once the attacker can reach the vulnerable service, exploitation is feasible. Organizations using Triton Inference Server in multi-tenant or cloud environments may face elevated risks due to shared infrastructure.

CWE-125Out-of-bounds Read (opens in a new tab)
Base weaknessCWE Top 25 (2025): #8
The product reads data past the end, or before the beginning, of the intended buffer.
Implementation · Input Validation
Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does.
Architecture and Design · Language Selection
Use a language that provides appropriate memory abstractions.
OWASP Cheat Sheets
XML Security (opens in a new tab)
Mass Assignment (opens in a new tab)
OWASP ASVS 4.0.3
V13.3.1 (opens in a new tab)
V5.1.4 (opens in a new tab)
V5.1.3 (opens in a new tab)
Attack patterns
Overread Buffers (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).

References

Community reviews

Loading reviews…

Exploitation evidence

What to do

Affected software

Description

AI analysisMachine-generated, updated 6 Aug 2025, 13:19 UTC

WeaknessesCWE-125

Technical details

References

Community reviews

Related threats