Critical Axios Vulnerability Enables Remote Code Execution, PoC Released

A critical security vulnerability has been discovered in Axios, one of the most widely used HTTP client libraries, exposing applications to Remote Code Execution (RCE) and full cloud infrastructure compromise.

Tracked as CVE-2026-40175, this flaw carries a critical CVSS 3.1 score of 9.9 and allows attackers to bypass AWS IMDSv2 security controls to exfiltrate sensitive cloud metadata.

Following its discovery by security researcher raulvdv and subsequent publication by jasonsaayman, a public Proof of Concept (PoC) has been released.

The Gadget Attack Chain

The vulnerability is rooted in an unrestricted cloud metadata exfiltration flaw via a header injection chain within the axios npm package.

According to the advisory, the core issue resides in the library’s header processing components, specifically within the lib/adapters/http.js file.

The defect involves a complete lack of HTTP header sanitization (CWE-113), which, when combined with default Server-Side Request Forgery (SSRF) capabilities (CWE-918) and request smuggling (CWE-444), creates a highly exploitable attack vector.

What makes this vulnerability particularly dangerous is its requirement for zero direct user input. The exploit relies entirely on a “Gadget” chain.

If a threat actor successfully pollutes Object.prototype via any other third-party dependency in the application stack, such as body-parser, qs, or minimist, Axios will automatically merge these polluted properties during its configuration phase.

Because Axios fails to sanitize these merged values for carriage return and line feed (CRLF) characters, the property transforms into a devastating HTTP request smuggling payload.

The released PoC demonstrates a severe exploitation scenario targeting cloud environments. When a developer makes a completely safe, hardcoded outbound request, Axios inadvertently merges the polluted prototype property, such as x-amz-target, into the request headers.

These malicious headers are then written directly to the socket without any structural validation.

This behavior enables the attacker to smuggle a secondary PUT request targeting the AWS EC2 Metadata Service at 169.254.169.254.

Critically, this smuggled request includes the mandatory X-aws-ec2-metadata-token-ttl-seconds header, successfully bypassing AWS IMDSv2 session token protections.

The attacker can then retrieve a valid session token, steal IAM credentials, and achieve full cloud account compromise, alongside secondary impacts like authentication bypass and cache poisoning.

Affected Versions and Mitigation

The vulnerability affects a broad range of Axios deployments across both legacy and modern branches. All Axios versions prior to 1.13.2 are explicitly vulnerable to this gadget attack chain.

To mitigate this critical threat, development teams must immediately update their Axios dependencies to version 1.15.0 or later.

This patch resolves the issue by enforcing strict validation on all header values before passing them to the underlying request functions.

If the updated library detects any invalid CRLF characters, it will immediately throw a security error, effectively neutralizing the injection chain and preventing the request smuggling attack.

Follow us on Google News, LinkedIn, and X to Get Instant Updates and Set GBH as a Preferred Source in Google.