Software Supply Chain Attack on Axios HTTP

Client Alert

March 31, 2026

Norma

Dowler

Director of Product Marketing, Cyber

On March 31, 2026, a North Korea–linked threat actor conducted a software supply chain attack involving the widely used Axios JavaScript library ecosystem. Malicious code was introduced into npm packages designed to mimic or integrate with Axios, creating a pathway for credential theft and system compromise in downstream environments.

The campaign appeared selective rather than indiscriminate, with indicators suggesting that specific developers and organizations were targeted through poisoned dependencies rather than a broad compromise of the core Axios library itself.

What Happened

Google’s Threat Intelligence Group identified a cluster of malicious npm packages impersonating or closely resembling Axios-related modules, which were used to distribute obfuscated code capable of exfiltrating sensitive data and establishing persistence in infected environments (Google Cloud). These packages were published to npm with naming conventions designed to evade casual scrutiny, increasing the likelihood of accidental inclusion in development workflows.

Source: Google Threat Intelligence Group, Exiger Analysis. npm logo: Boboss74, Public domain, via Wikimedia Commons

According to reporting, the malicious packages were downloaded thousands of times before being removed, suggesting a non-trivial exposure window. Axios itself—one of the most widely used HTTP client libraries in JavaScript—was not directly compromised, but its ubiquity made it an effective lure.

Further analysis linked the activity to a North Korean threat actor consistent with prior campaigns focused on software developers and cryptocurrency-adjacent environments. The actor’s tradecraft aligns with earlier DPRK operations that leverage open source ecosystems to gain initial access into high-value targets.

How the Attack Worked

The attack leveraged a sophisticated dependency confusion–style approach combined with typosquatting and social engineering elements:

Package Impersonation: Threat actors published npm packages with names resembling legitimate Axios-related libraries, increasing the likelihood of developer adoption.
Malicious Payload Delivery: Once installed, the packages executed obfuscated scripts designed to collect environment variables, credentials, and system metadata.
Command-and-Control (C2): Exfiltrated data was transmitted to attacker-controlled infrastructure, enabling follow-on access.
Targeted Deployment: Evidence suggests selective targeting, potentially guided by reconnaissance of developer profiles or organizational affiliations.

This approach avoided direct tampering with Axios itself, instead exploiting trust in the broader ecosystem of packages that depend on or extend it.

Implications for Organizations

Developer-Centric Risk

Individual developers and small teams may serve as initial access points, especially where dependency vetting is informal.

Software Bill of Materials (SBOM) Gaps

Many organizations lack real-time visibility into transitive dependencies introduced during development.

Delayed Detection

Malicious packages can persist in registries long enough to achieve meaningful distribution before being flagged and removed.

Targeted Campaigns

The shift from broad to selective targeting increases the likelihood that traditional anomaly detection may miss early-stage compromise.

Why This Matters

This incident reinforces a persistent structural risk in open source ecosystems: trust is often transitive, and widely adopted libraries create gravitational pull for malicious lookalikes. Even when a core package remains secure, adjacent dependencies can introduce compromise pathways that are difficult to detect through traditional controls.

Attribution to a North Korean threat actor is particularly significant. DPRK-linked groups have consistently targeted developers, cryptocurrency platforms, and technology providers as part of broader revenue-generation and espionage campaigns. The selective nature of this operation suggests an evolution toward more precise targeting within developer ecosystems.

For organizations, the incident highlights the limitations of perimeter-focused security models when software supply chains are inherently decentralized and continuously changing.

Exiger Recommendations

Organizations should treat this incident as a prompt to reassess software supply chain controls, particularly within developer environments:

➔ Strengthen Dependency Governance

Implement strict allowlisting and verification processes for third-party and open source packages, including namespace validation and publisher reputation checks.

For example, the Exiger Cyber Dependency Version Locking check would have blocked dependency files using Axios without a locked version and prevented proliferation of the malicious version.

➔ Enhance Visibility into Dependencies

Maintain continuously updated SBOMs that include transitive dependencies and integrate them into risk monitoring workflows.

➔ Monitor Developer Environments

Extend security telemetry and anomaly detection into developer workstations and CI/CD pipelines, where initial compromise is most likely to occur.

➔ Leverage Threat Intelligence

Incorporate real-time intelligence on malicious packages and emerging threat actor tactics to reduce detection lag.

➔ Simulate Supply Chain Attacks

Regularly test organizational resilience to dependency-based compromise scenarios to identify detection and response gaps.

Exiger Cyber maps software dependencies to known threat activity and provide continuous monitoring of open source ecosystems to materially reduce exposure to this class of attack.

Get in touch to request a demo and see how Exiger Cyber can help you get out in front of cyber risks.

More Resources: