/

CVE-2026-24170

Severity:

HIGH RISK

(8.8)

Improper Authentication (CWE-287) in NVIDIA UFM Web UI

CVE-2026-24170

Undergoing Analysis

Vulnerability type

Category:

Improper Authentication

CWE:

CWE-287

Impact:

Code Execution and Privilege Escalation

Severity:

HIGH RISK

(8.8)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

NVIDIA UFM Enterprise contains an improper authentication vulnerability in its web interface authorization component. An authenticated user who sends specially crafted HTTP requests may bypass intended authentication checks, potentially enabling code execution and privilege escalation.

Affected context

The vulnerability occurs in the NVIDIA UFM Enterprise web interface when authorization-related HTTP requests are processed without consistently enforcing the intended authentication state or access controls. Exploitation generally depends on access to the web interface through an authenticated account and on the affected authorization path being reachable.

Exploit preconditions

  • Attack surface: NVIDIA UFM Enterprise web interface
  • Required access: An authenticated user account with access to the relevant HTTP endpoints
  • Trigger point: Processing of specially crafted HTTP requests by the authorization component
  • Why exploitation is possible: Authentication or authorization checks may be improperly applied, allowing a request to reach privileged functionality beyond the user's intended permissions

Why this happens

The underlying bug is an authentication enforcement failure in a web authorization component. When request handling does not reliably bind sensitive operations to a validated user identity and authorization context, an authenticated user may influence how access checks are evaluated. This behavior maps to CWE-287: Improper Authentication, where the application does not correctly establish or enforce the identity required for protected operations.

Potential impact

  • An attacker may execute code in the affected application environment
  • An attacker could escalate privileges within the UFM deployment
  • Sensitive operational or infrastructure data may be accessed or modified
  • Administrative functionality could be abused to affect managed systems
  • The vulnerability may increase the risk of lateral movement from the management environment

General mitigation guidance

Mitigation typically focuses on ensuring that every protected operation is evaluated against a strongly validated identity and explicit authorization policy.

Common defensive patterns include:

  • Centralizing authentication and authorization checks for sensitive web actions
  • Enforcing server-side authorization independently of client-controlled request values
  • Applying deny-by-default access policies and least-privilege roles
  • Binding sessions, identities, and privileges to well-defined server-side context
  • Limiting exposure of the management interface to trusted administrative networks
  • Requiring strong authentication and monitoring unusual authenticated activity

How teams detect issues like this

Organizations commonly detect improper authentication risks through:

  • Static analysis for endpoints that perform privileged actions without a shared authorization guard
  • Code review checks for authentication decisions based on client-controlled parameters or incomplete session state
  • Dynamic testing of authenticated endpoints with users holding different privilege levels
  • Monitoring for unusual requests, privilege transitions, and administrative actions from non-administrative accounts
  • Policy enforcement requiring centralized access-control middleware for management interfaces

Automated security review tools may help identify inconsistent authentication and authorization patterns, but findings should be validated against application-specific access policies.

Sources

SourceLink
github.comgithub.com/NVIDIA/product-security/tree/main/2026/5809
nvd.nist.govnvd.nist.gov/vuln/detail/CVE-2026-24170
cve.orgcve.org/CVERecord?id=CVE-2026-24170

Frequently asked

questions

It may be exploitable over the network when the NVIDIA UFM web interface is reachable and the attacker has a valid authenticated account. The practical scope depends on network exposure, account permissions, and endpoint access controls.

Exploitation generally requires authentication to the web interface and the ability to send requests to the affected authorization path. A highly privileged account may not be necessary if authorization checks can be bypassed.

Ensure that sensitive operations use consistent server-side authentication and authorization checks, with explicit identity validation and least-privilege enforcement for every protected endpoint.

Use centralized access-control middleware, test endpoints across multiple privilege levels, reject client-controlled authorization decisions, and monitor management interfaces for anomalous authenticated activity.

Find out what's already
exploitable in your codebase.