5.3 CVE-2025-6920

CSRF

A flaw was found in the authentication enforcement mechanism of a model inference API in ai-inference-server. All /v1/* endpoints are expected to enforce API key validation. However, the POST /invocations endpoint failed to do so, resulting in an authentication bypass. This vulnerability allows unauthorized users to access the same inference features available on protected endpoints, potentially exposing sensitive functionality or allowing unintended access to backend resources.
https://nvd.nist.gov/vuln/detail/CVE-2025-6920

Categories

CWE-306 : Missing Authentication for Critical Function
The product does not perform any authentication for functionality that requires a provable user identity or consumes a significant amount of resources. For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks entirely. Then, these modified values would be submitted to the server. When storing data in the cloud (e.g., S3 buckets, Azure blobs, Google Cloud Storage, etc.), use the provider's controls to require strong authentication for users who should be allowed to access the data [REF-1297] [REF-1298] [REF-1302]. File-sharing PHP product does not check if user is logged in during requests for PHP library files under an includes/ directory, allowing configuration changes, code execution, and other impacts. Chain: a digital asset management program has an undisclosed backdoor in the legacy version of a PHP script (CWE-912) that could allow an unauthenticated user to export metadata (CWE-306) TCP-based protocol in Programmable Logic Controller (PLC) has no authentication. Condition Monitor firmware uses a protocol that does not require authentication. SCADA-based protocol for bridging WAN and LAN traffic has no authentication. Safety Instrumented System uses proprietary TCP protocols with no authentication. Distributed Control System (DCS) uses a protocol that has no authentication. Chain: Cloud computing virtualization platform does not require authentication for upload of a tar format file (CWE-306), then uses .. path traversal sequences (CWE-23) in the file to access unexpected files, as exploited in the wild per CISA KEV. Bluetooth speaker does not require authentication for the debug functionality on the UART port, allowing root shell access WiFi router does not require authentication for its UART port, allowing adversaries with physical access to execute commands as root IT management product does not perform authentication for some REST API requests, as exploited in the wild per CISA KEV. Default setting in workflow management product allows all API requests without authentication, as exploited in the wild per CISA KEV. MFV. Access TFTP server without authentication and obtain configuration file with sensitive plaintext information. Agent software running at privileges does not authenticate incoming requests over an unprotected channel, allowing a Shatter" attack. Product enforces restrictions through a GUI but not through privileged APIs. monitor device allows access to physical UART debug port without authentication Programmable Logic Controller (PLC) does not have an authentication feature on its communication protocols.

References

secalert@redhat.com

https://access.redhat.com/security/cve/CVE-2025-6920 Vendor Advisory
https://bugzilla.redhat.com/show_bug.cgi?id=2375522 Issue Tracking Vendor Advisory

CPE

cpe	start	end
Configuration 1
cpe:2.3:a:redhat:ai_inference_server:-:::::::*

REMEDIATION

EXPLOITS

Exploit-db.com

id	description	date
No known exploits

POC Github

Url
No known exploits

Other Nist (github, ...)

Url
No known exploits

CAPEC

Common Attack Pattern Enumerations and Classifications

id	description	severity
12	Choosing Message Identifier This pattern of attack is defined by the selection of messages distributed via multicast or public information channels that are intended for another client by determining the parameter value assigned to that client. This attack allows the adversary to gain access to potentially privileged information, and to possibly perpetrate other attacks through the distribution means by impersonation. If the channel/message being manipulated is an input rather than output mechanism for the system, (such as a command bus), this style of attack could be used to change the adversary's identifier to more a privileged one. [Determine Nature of Messages] Determine the nature of messages being transported as well as the identifiers to be used as part of the attack [Authenticate] If required, authenticate to the distribution channel [Identify Known Client Identifiers] If any particular client's information is available through a control channel available to all users, the adversary will discover particular identifiers for targeted clients by observing this channel, or requesting client information through this channel. [Change Message Identifier] Adversaries with client access connecting to output channels could change their channel identifier and see someone else's (perhaps more privileged) data.	High
166	Force the System to Reset Values An attacker forces the target into a previous state in order to leverage potential weaknesses in the target dependent upon a prior configuration or state-dependent factors. Even in cases where an attacker may not be able to directly control the configuration of the targeted application, they may be able to reset the configuration to a prior state since many applications implement reset functions.	Medium
216	Communication Channel Manipulation An adversary manipulates a setting or parameter on communications channel in order to compromise its security. This can result in information exposure, insertion/removal of information from the communications stream, and/or potentially system compromise.
36	Using Unpublished Interfaces or Functionality An adversary searches for and invokes interfaces or functionality that the target system designers did not intend to be publicly available. If interfaces fail to authenticate requests, the attacker may be able to invoke functionality they are not authorized for. [Identify services] Discover a service of interest by exploring service registry listings or by connecting on a known port or some similar means. [Authenticate to service] Authenticate to the service, if required, in order to explore it. [Identify all interfaces] Determine the exposed interfaces by querying the registry as well as probably sniffing to expose interfaces that are not explicitly listed. [Attempt to discover unpublished functions] Using manual or automated means, discover unpublished or undocumented functions exposed by the service. [Exploit unpublished functions] Using information determined via experimentation, exploit the unpublished features of the service.	High
62	Cross Site Request Forgery An attacker crafts malicious web links and distributes them (via web pages, email, etc.), typically in a targeted manner, hoping to induce users to click on the link and execute the malicious action against some third-party application. If successful, the action embedded in the malicious link will be processed and accepted by the targeted application with the users' privilege level. This type of attack leverages the persistence and implicit trust placed in user session cookies by many web applications today. In such an architecture, once the user authenticates to an application and a session cookie is created on the user's system, all following transactions for that session are authenticated using that cookie including potential actions initiated by an attacker and simply "riding" the existing session cookie. [Explore target website] The attacker first explores the target website to determine pieces of functionality that are of interest to them (e.g. money transfers). The attacker will need a legitimate user account on the target website. It would help to have two accounts. [Create a link that when clicked on, will execute the interesting functionality.] The attacker needs to create a link that will execute some interesting functionality such as transfer money, change a password, etc. [Convince user to click on link] Finally, the attacker needs to convince a user that is logged into the target website to click on a link to execute the CSRF attack.	Very High

Cybersecurity needs ?

Strengthen software security from the outset with our DevSecOps expertise

Integrate security right from the start of the software development cycle for more robust applications and greater customer confidence.
Our team of DevSecOps experts can help you secure your APIs, data pipelines, CI/CD chains, Docker containers and Kubernetes deployments.

Discover this offer

Soutenez No Hack Me sur Tipeee