North Korean-linked cybercriminals have been identified as the perpetrators behind a series of attacks leveraging trusted npm packages to initiate extensive software supply-chain breaches. These sophisticated operations involve compromising legitimate package maintainers’ accounts, allowing them to insert malicious updates into software tools widely trusted by developers and enterprises.
Targeted npm Packages in Cyber Campaigns
The attackers have successfully infiltrated packages such as typo-crypto, debug, chalk, and axios, over a span from March 2025 to March 2026. Organizations that routinely update to the latest versions risk inadvertently installing harmful code, which can seamlessly integrate into their ongoing development and build operations.
Security experts at AWS have attributed these attacks to a North Korean-linked group identified as SAPPHIRE SLEET, STARDUST CHOLLIMA, BlueNoroff, CageyChameleon, and Alluring Pisces. In a detailed report shared with Cyber Security News, AWS highlighted the group’s use of similar techniques, malicious npm packages, and repeated code patterns across these incidents.
Impact on Open Source Dependencies
The ramifications of these attacks are significant due to the widespread use of open-source dependencies in numerous applications, cloud environments, and automated build systems. A single compromised update can expose multiple downstream environments simultaneously, echoing the vulnerabilities seen in other npm supply-chain compromises.
Social engineering played a crucial role, as the attackers manipulated trusted maintainer accounts to distribute updates embedded with concealed malicious code, thus bypassing the scrutiny typically applied to unfamiliar dependencies. Among the largest affected was the axios package, with over 100 million weekly downloads at the time of compromise in March 2026.
Evolution of Attack Strategies
AWS analysts warned that the attackers are increasingly distributing harmful activities across multiple ordinary-looking packages. This strategy makes isolated reviews less effective as one package might store encrypted data, another might unlock it, and a third could retrieve the final payload.
Additionally, threat actors are building trust by maintaining legitimate projects or contributing to existing ones before exploiting their position. This patient approach mimics the tactics seen in the XZ Utils incident and recent North Korean npm attacks, where legitimate access becomes a critical asset for attackers.
Generative AI is further complicating the threat landscape by enabling attackers to create realistic documentation and code, potentially misleading developers and coding assistants into installing malicious packages.
Strategies for Mitigating Risks
Security teams are advised to maintain a comprehensive inventory of dependencies, scrutinize unexpected package changes, and monitor package behavior during installation and runtime. Limiting automatic updates and investigating anomalous network activities from build systems or developer workstations are also recommended practices.
In the event of a suspected dependency compromise, organizations should treat it as a broader credential and environment risk. This involves reviewing affected repositories, isolating compromised systems, rotating credentials when necessary, and monitoring for subsequent malicious activities to prevent a package compromise from escalating into a larger security breach.
