Open source malware refers to malicious code hidden within software packages on platforms like npm, PyPI, and Hugging Face, which are increasingly targeted by cybercriminals. In the third quarter of this year, Sonatype identified 34,319 malicious open source packages, contributing to a total of 877,522 over the past six years. Attackers are using artificial intelligence to embed malware in tools developers rely on, focusing on stealthy operations to avoid detection. Data exfiltration malware made up 37% of detected malicious packages in Q3, targeting sensitive information. Nearly 38% of threats were categorized as “droppers,” which install additional malware. The prevalence of backdoor-laden packages increased by 143% from the previous quarter. Financial organizations were the primary targets, with 47% of blocked malware attempts aimed at banks and financial services. Notable incidents include the npm hijacking of the “chalk” and “debug” packages and the Shai-Hulud campaign, which spread automatically across repositories. In the first quarter of 2025, 17,954 malicious packages were reported, with over half targeting confidential data, indicating a shift towards organized attacks on trusted dependencies.