A recent report by Amazon has flagged a surge in software supply chain attacks, attributing a series of compromises in open source libraries to a North Korea-backed threat group. The findings highlight growing cyber risks for developers, driven by the misuse of generative AI, targeted attacks on code repositories, and financially motivated operations by nation-state actors.
The group—tracked under multiple names including Sapphire Sleet, Stardust Chollima, BlueNoroff, CageyChameleon, and Alluring Pisces—has been linked to several attacks involving the insertion of malicious code into packages hosted on the npm repository over the past few years.
“When an attacker compromises a widely used open source package, every organization that depends on that package is potentially affected,” said CJ Moses, Chief Information Security Officer at Amazon Integrated Security. He added that both the volume and sophistication of such attacks have increased significantly, largely driven by DPRK-linked actors and cybercriminal groups.
Supply chain attacks are gaining traction as they allow threat actors to infiltrate thousands of downstream systems by targeting a limited number of widely used packages, rather than attacking organisations individually.
Widening Threat Landscape
Other cybersecurity researchers have reported similar trends. Socket researchers earlier this year identified North Korea-linked actors deploying dozens of malicious npm packages to steal developer credentials. GitHub also revealed that it has been banning an average of 11 accounts per month in 2025 for distributing malware tied to these groups.
A study by a researcher at Pusan National University described North Korea’s cyber operations as a core element of its national strategy, used to offset economic constraints and geopolitical isolation through cybercrime and espionage.
Pattern of Package Compromises
Amazon’s report details a series of incidents in 2025, including the compromise of the typo-crypto package in March, followed by attacks on debug and chalk packages in September, and the widely used axios library, which records over 100 million weekly downloads.
Researchers have, for the first time, linked these attacks to a single North Korean group. The modus operandi typically involves social engineering tactics to gain access to package maintainers’ credentials, followed by the release of infected updates. Organisations that automatically integrate these updates unknowingly introduce malicious code into their systems.
Security firms including Microsoft, Google Threat Intelligence Group, and Wiz have independently attributed these campaigns to DPRK-linked actors, noting widespread impact across cloud environments.
Evolving Attack Techniques
The report also underscores how attackers are refining their methods. Tactics now include distributing malicious workflows across multiple seemingly benign packages, building long-term trust by maintaining legitimate projects before injecting malware, and deploying multi-stage payloads secured with advanced cryptographic techniques.
Generative AI is further complicating detection. According to Amazon, AI-generated malicious code increasingly mimics legitimate development patterns, eliminating traditional red flags such as poor documentation or inconsistent coding styles.
Developers in the Crosshairs
Developers have emerged as high-value targets due to their access to sensitive credentials, source code, and deployment pipelines. The rise of AI-assisted coding tools is expanding this risk surface, as more users and systems interact directly with third-party code and repositories.
Open source platforms remain particularly vulnerable due to their collaborative nature, automatic update mechanisms, and high levels of trust within developer communities.
The report concludes that as software ecosystems become more interconnected, securing the open source supply chain will be critical to mitigating large-scale cyber threats.
