External risk intelligence

Network-AI ApprovalInbox Unauthorized Access Vulnerability.

CVE advisorySeverity: CRITICAL (CVSS 9.3)

CVE-2026-64622

The vulnerability affects web-based GET routes within an application that provides approval and workflow management functionality. Such interfaces, including API endpoints and approval inboxes, are commonly deployed as network-accessible web services for distributed teams, making them frequently reachable in standard web application environments.

Halo Surface Signal: 4 out of 5 — likely to be public-facing.

External exposure likelihood

Horizon Alert

Summary of the vulnerability and why it matters

A recently identified vulnerability in Network-AI software allows unauthenticated access to sensitive approval request details, including command strings and risk levels. This issue, stemming from a failure to properly enforce authorization checks, could expose operational information to unauthorized actors. The main concern is confirming relevance and exposure to sensitive approval data.

Unauthorized access to approval details is possible.
Protects sensitive operational and command information.
Assess exposure to sensitive approval data.

Attack Path

How an attacker could exploit the issue

An attacker could access sensitive approval request details, even without authentication, by exploiting a flaw in how the application handles authorization checks on certain network-accessible routes. This could expose details such as shell commands, file paths, and justifications.

No prior authentication required.
Accessing specific network routes.
Sensitive data disclosure.

Live Threat

Current exploitation, exposure, and threat context

This vulnerability could expose sensitive details about approval requests, including commands to be executed, file paths, and risk assessments, to unauthenticated actors. These details could be disclosed from any website an operator visits when the affected application is running.

Approval request data and commands.
Unauthenticated access to GET routes.
Sensitive information disclosure to any website.

Operational Fix

Recommended remediation, mitigation, and detection steps

The Network-AI application's authorization flaw requires a joint effort between application and platform teams to address. The initial step is to pinpoint all Network-AI deployments, ascertain their external reachability and business criticality, and identify the designated owner for each instance. This information will then inform a prioritized remediation plan.

Application owners should manage this issue.
Verify external reachability and business criticality.
Plan remediation based on identified risk.

Supplementary metadata

CVSS vector

Validate whether this threat affects your internet-facing exposure.

Halo Threat Intelligence helps prioritize remediation with Halo Surface Signal and H/A/L/O context. Start exposure validation with a free external attack surface trial.

Free EASM Trial See How Signal Works

Frequently asked questions

What is the Network-AI software component?

Network-AI is an npm package used by developers to manage workflows and internal approval processes. It provides an 'ApprovalInbox' feature that allows teams to track, review, and authorize operational requests, such as command execution strings, file access justifications, and associated risk levels.

What is the vulnerability in CVE-2026-64622?

This vulnerability is classified as CWE-862, which is a Missing Authorization weakness. Essentially, the software fails to enforce the security secrets or authentication checks that administrators have configured. This allows anyone to view sensitive records in the ApprovalInbox without proving their identity.

How does an attacker trigger this bug?

An attacker can trigger this by simply navigating to specific GET routes within the ApprovalInbox, such as those used to view approval lists or individual status requests. The bug occurs regardless of the authentication settings; therefore, sending a standard request to these endpoints is sufficient to bypass security and retrieve the data. Authentication is not required for the exploit to succeed.

Why should I care about this vulnerability?

According to Halo Surface Signal, this software is often deployed as a web service accessible to distributed teams, meaning it is frequently exposed to network traffic. Because the API endpoints are reachable over the network, your ApprovalInbox could be visible to unauthenticated actors if it is accessible via the internet or an untrusted network. Additionally, the cross-origin header makes it possible for other websites to access this data if an operator visits them.

What are the first steps to address this?

Start by identifying every location where Network-AI is currently deployed in your environment. Prioritize these instances based on their business sensitivity and network reachability. Once you have a complete inventory, work with the relevant application owners to plan and coordinate the necessary remediation updates.

References

Written by Nicholas Merritt

Nicholas Merritt is VP of Security Solutions at Halo Security. With more than 25 years across application security, vulnerability prioritization, and offensive security, he helps teams translate threat data into practical exposure validation and remediation focus. He authors Halo Threat Intelligence CVE advisories using Halo Surface Signal and H/A/L/O context to prioritize internet-facing risk.