Amazon Funds Akrites Initiative After Sapphire Sleet NPM Attacks

Security researchers and cloud computing giants are confronting a sophisticated, long-term open-source software campaign targeting the digital supply chain. Amazon connected multiple high-profile compromises within the Node Package Manager ecosystem to the Sapphire Sleet threat actor, also recognized as BlueNoroff and Stardust Chollima.

Attackers then escalated their efforts in September 2025 by compromising the widely used debug and chalk libraries, which managed to reach an estimated 10% of cloud environments within two hours of publication. By March 2026, the targeting shifted to axios, one of the most popular npm packages boasting over 100 million weekly downloads.

Social Engineering Maintainers and Weaponizing AI

The mechanics of the npm intrusions relied heavily on human manipulation rather than traditional zero-day exploits. The attacker gained access by socially engineering package maintainers to publish malicious updates that automatically rolled out to downstream developers. Investigators assigned a medium confidence level to the Sapphire Sleet attribution, citing shared tactics, techniques, and procedures, overlapping command-and-control infrastructure, and consistent operational patterns.

Beyond standard social engineering, the campaign highlighted emerging threats driven by artificial intelligence. Threat actors spent months building credibility by maintaining legitimate code or contributing to projects before injecting malicious code. They split payloads across benign packages, relied on external scripts for weaponization, and utilized environment-aware malware that delayed execution unless real developer or production sandboxes were detected. Attackers also exploited slopsquatting by registering package names hallucinated by AI coding assistants, betting that developers or autonomous coding agents would automatically install them.

Parallel Crypto Attacks and State-Level Revenue Streams

While software supply chains faced automated and AI-assisted infiltration, decentralized finance platforms absorbed parallel shocks from targeted phishing operations. Humanity Protocol suffered a $36 million theft of Humanity (H) tokens after a compromised employee laptop allowed attackers to gain full remote access as detailed in Quantstamp incident response findings.

Kim Jong Un Accuses Japan of Transforming Into a War State

Blockchain security firm Quantstamp identified a malicious attachment disguised as a token lockup schedule update from South Korean cryptocurrency exchange Bithumb. The malware was signed with a South Korean Hancom digital certificate—a signature pattern described by security analysts as characteristic of DPRK intrusions—which enabled attackers to copy project director Chong Yee Wai’s MetaMask wallet credentials and private keys.

These incidents fit into a much larger, industrialized financial apparatus. Over the past decade, North Korea-linked actors stole an estimated $6.75 billion in cryptocurrency across 263 documented incidents, according to a report from CertiK. Those figures included at least $578 million tied to North Korean threat actors out of $634 million stolen in crypto exploits during April alone.

Internal Financial Fallout and Industry Defense Responses

“The ring converted stolen state trade funds into cryptocurrency and smuggled large amounts of foreign currency in border regions, the source said… The breach struck at the core of North Korea’s financial system. It came from within, carried out by the country’s own IT workforce.”

Anonymous source, via Daily NK

News of the internal scandal generated widespread alarm among high-ranking military and academic circles. Meanwhile, state media outlets continued to deny involvement in global cybercrime, with a Foreign Ministry spokesperson featured in the Korean Central News Agency dismissing allegations as a non-existent ‘cyber threat’ pushed by the United States.

In response to the expanding scope of supply-chain compromises, Amazon deployed a multi-faceted defense strategy by investing $12.5M in the Akrites initiative, which collaborates with OpenSSF and industry partners to protect critical open-source software against AI-enabled attacks.